Biomedical Engineering for Global Health (Cambridge Texts in Biomedical Engineering)

  • 87 1,260 10
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

Biomedical Engineering for Global Health (Cambridge Texts in Biomedical Engineering)

This page intentionally left blank Biomedical Engineering for Global Health Rebecca Richards-Kortum’s pioneering work

5,165 207 19MB

Pages 418 Page size 235 x 327 pts Year 2009

Report DMCA / Copyright


Recommend Papers

File loading please wait...
Citation preview

This page intentionally left blank

Biomedical Engineering for Global Health Rebecca Richards-Kortum’s pioneering work and teaching are captured in this motivating text written for non-science majors and biomedical engineers, inspiring them to engage in solving the global health issues that face us all. Studying with Biomedical Engineering for Global Health, students will: r begin to understand the medical, regulatory, economic, social, and ethical challenges facing the development of health systems; r see how these constraints affect the design of new devices and therapies; r learn through case studies, including cancer screening, imaging technologies, implantable devices, and vaccines; and r read profiles of undergraduate students who have participated in international technology development internships in Africa. Rebecca Richards-Kortum is the Stanley C. Moore Professor of Bioengineering at Rice University, where her research group is currently developing miniature microscopes and low-cost imaging systems to enable early detection of pre-cancerous changes in living tissue. In collaboration with colleagues at Rice University and the Texas Medical Center, she has developed a four-year multi-disciplinary education and training program that promotes engineering and engineering technology on a global scale. Prior to working at Rice University, Dr. Richards-Kortum was a Professor of Biomedical Engineering at the University of Texas at Austin, where she was elected to the Academy of Distinguished Teachers and received the Chancellor’s Council Outstanding Teaching Award. She has also received many other awards, including being named a Piper Professor for excellence in teaching by the Minnie Stevens Piper Foundation in 2004, and receiving the Sharon Keillor Award for Women in Engineering (2004) and the Chester F. Carlson Award (2007) from the American Society for Engineering Education. In 2008, she was elected to the National Academy of Engineering. “This beautifully written volume by Rebecca Richards-Kortum will inspire and empower the next generation of engineers to make global health their calling.” Thomas Kalil, UC Berkeley and Clinton Global Initiative “This book will become the most influential biomedical text of our generation . . . No other book has made such a transformation of young scientists in such a short time.” Nicholas Peppas, The University of Texas at Austin “This book is an excellent first step in educating engineers about medical problems in the developing worlds and ways in which bioengineers can make a difference.” Paul Yager, University of Washington, Seattle

Cambridge Texts In Biomedical Engineering Series Editors W. Mark Saltzman, Yale University Shu Chien, University of California, San Diego Series Advisors William Hendee, Medical College of Wisconsin Roger Kamm, Massachusetts Institute of Technology Robert Malkin, Duke University Alison Noble, Oxford University Bernhard Palsson, University of California, San Diego Nicholas Peppas, University of Texas at Austin Michael Sefton, University of Toronto George Truskey, Duke University Cheng Zhu, Georgia Institute of Technology Cambridge Texts in Biomedical Engineering provides a forum for high-quality accessible textbooks targeted at undergraduate and graduate courses in biomedical engineering. It covers a broad range of biomedical engineering topics from introductory texts to advanced topics including, but not limited to, biomechanics, physiology, biomedical instrumentation, imaging, signals and systems, cell engineering, and bioinformatics. The series blends theory and practice, aimed primarily at biomedical engineering students, it is also suitable for broader courses in engineering, the life sciences and medicine.

Biomedical Engineering for Global Health Rebecca Richards-Kortum Rice University, Houston, Texas


Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York Information on this title: © R. Richards-Kortum 2010 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2009 ISBN-13


eBook (NetLibrary) (with enhanced colour)




Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.



page vii

1 Emerging medical technologies: high stakes science and the need for technology assessment 2 Bioengineering and technology assessment

1 21 31

4 World health and global health challenges


5 Healthcare systems: a global comparison


7 The evolution of technology: scientific method, engineering design, and translational research


9 Ethics of clinical research


10 Technologies for early detection and prevention of cancer


11 Cost effectiveness of screening for disease 305

3 Health and economic data: a global comparison

6 Healthcare costs vs. time: trends and drivers

8 Prevention of infectious disease



12 Technologies for treatment of heart disease


13 Clinical trial design and sample size calculation


14 Technology diffusion


15 Regulation of healthcare technologies


16 Future of bioengineering and world health




Dedication To my mostly patient children, Alex, Max, Zach and Kate, and to their endlessly patient father, Phil.

Acknowledgments I have been blessed to work with a wonderful team of creative, dedicated, and inspiring colleagues. Without their tireless efforts, this project would never have reached completion. Their contributions are most gratefully acknowledged. My thanks to Deanna Buckley, Christine Edwards, Dan Erchick, Jessica Gerber, Allison Lipper, Yvette Mirabal, Guadalupe Rodriguez, Richard Schwarz, Lauren Vestewig, and Rachel Wergin. Finally, my thanks to the many BME301, BIOE 301 and BTB students who provided thoughtful feedback and inspiration along the way.


Biomedical Engineering for Global Health gives students a cohesive overview of how biomedical technolo-

2. Who pays to solve problems in healthcare and how does this vary throughout the world?

gies are developed and translated into clinical practice. The text integrates the major diseases facing developed and developing countries with the recent technological advances and the economic, social, ethical and regula-

3. How can we use technology to solve world health problems? 4. How do new technologies move from the laboratory to the bedside?

tory constraints which impact the development of new technologies. Biomedical Engineering for Global Health is accessible to students from all disciplines. The text responds to student interest in the fields of bioengineering and global health. As the world becomes more interconnected, students seek more opportunities to learn about disease and health, and how science and engineering can be used to solve global health challenges. In a global context, the text introduces students to bioengineering, epidemiology, health disparities, and the development of medical drugs and devices. For introductory courses in bioengineering, global health, epidemiology or related fields, this text serves as a comprehensive overview of global health challenges and the methods to improve health and prevent disease. The text answers four primary questions. 1. What are the major health problems worldwide and how do these differ throughout the world?

Throughout the text, three major case studies are used to illustrate the development, assessment and global diffusion of new medical technologies, including development of new vaccines to prevent infectious disease, development of imaging technologies to improve early cancer screening and development of implantable devices to treat heart disease. The case studies and other examples help students understand the economic challenges associated with developing health systems. Frequent examples are used to contrast health systems in both developed and developing countries. The text includes profiles of leaders in translational research to expose students to the variety of career paths taken by individuals with MD, PhD or MD/PhD degrees. Also included are profiles of undergraduate students who have participated in international technology development internships in four countries in Africa. Students can directly relate what they are learning in the text to the experiences of their peers. These profiles will help



young bioengineering or global health students understand the important aspects of their discipline in the context in which it is practiced.

To further engage students in real world problems, a series of interactive classroom activities have been developed to accompany the lecture materials for the

Homework problems engage a broad audience in mathematical and graphical analysis of real biomedical data, as well as in writing about the social implications of technology development. In addition, a project assignment spans the text, guiding students through the design of a clinical trial to test a new technology. The project provides an opportunity for students to develop, expand and test their knowledge of sub-

course. These activities contextualize real global health problems so that students can better understand and begin to view problems and solutions simultaneously. Multi-media materials and connections to new accounts of scientific developments increase student engagement. Instead of simply focusing on the study of science and technology, this text takes an engaging studentcentered, contextual approach to the study of bioengi-

ject matter in a global context. The project asks students to select a disease of global health significance that is of interest to them. Students research current medical technologies to diagnose or treat the disease, and the limitations of those technologies in a resource-

neering and biotechnology. Unlike other similar texts, Biomedical Engineering for Global Health is designed for students from all disciplines. It places a strong emphasis on the need for new health technologies, the process of technology development and the impact of technol-

constrained setting. Design constraints are outlined for

ogy development in a personalized, global perspective.

a new technology to operate in a resource-limited setting. Finally students propose a new medical technology to diagnose or treat the disease which meets these constraints and design a clinical trial to test the

Understanding these processes is vitally important to students throughout their lives as they make decisions about their own medical care and contribute to discussion of public policy issues affecting healthcare through-


out the world.

1 Emerging medical technologies: high stakes science and the need for technology assessment

In the past century, advances in medical technology have yielded enormous improvements in human health. For example, our scientific understanding of the immune response and the resulting development of vaccines has vastly reduced the incidence of many infectious diseases. Smallpox has killed more people throughout history than perhaps any other infectious disease. Yet, in 1980, the World Health Organization announced that smallpox had been eradicated worldwide through a program of vaccination (Figure 1.1). Despite these advances, many medical technologies are available to only a small segment of the world’s population that can afford them. Today, emerging technologies have the potential to transform the future of healthcare, offering the potential to diagnose and prevent disease before it strikes, to treat disease in a targeted manner, and to utilize cells and genes for patient-specific therapies. For example, gene therapy offers the promise to cure fatal genetic diseases such as cystic fibrosis and to reprogram a patient’s immune system to more effectively fight HIV/AIDS, the leading cause of death in sub-Saharan Africa. Sequencing the genome of M. tuberculosis has pointed to new molecular targets for more effective drugs to treat tuberculosis. Small silicon chips containing every gene in the human genome may soon be used to detect cancer at the earliest and most curable stages and to

Figure 1.1. The development of the smallpox vaccine and the subsequent eradication of the disease is an example of a powerful medical technology. CDC/ World Health Organization; Stanley O. Foster.


Biomedical Engineering for Global Health

individually tailor therapeutic agents for each patient. Tissue engineering holds the promise to create artificial organs, overcoming problems with the limited supply of donor organs. Novel, biologically active materials may be used to coat blood vessels within the heart to prevent heart attacks, one of the leading causes of death in the United States.

Medical technology The use of novel technologies to develop new drugs, biologics, or medical devices designed to diagnose, treat or prevent disease. Bioengineering The application of engineering design to develop new medical technologies.

which can significantly impact world health. Throughout the book, we present and apply tools to systematically evaluate these new medical technologies. The book is organized to address four central questions.

Four central questions addressed (1) What are the major human health problems worldwide and how do these differ throughout the world? (2) Who pays to solve problems in healthcare and how does this vary throughout the world? (3) How can we use technology to solve world health problems? (4) How do new technologies move from the laboratory to the bedside?

Biotechnology The use of living systems to make or improve new products, frequently targeted toward improving human health.

What is needed to bring these new technologies from the research laboratory to your physician’s office in a safe and affordable way? As a society, how should we invest our limited financial and human resources to develop new medical technologies? Can new technologies reduce global disparities in health or will they simply widen the gap in health status between developing and developed countries? In this textbook, we examine how bioengineers integrate advances in the physical, information and life sciences to develop new medical technologies. To be effective, new healthcare technologies must provide a better means of preventing, detecting or treating disease. At the same time, technologies must also be affordable to those who need them. The goal of bioengineering is to harness science to solve health problems in the face of such constraints. Our study of bioengineering for world health is organized to first understand both global health needs and resource limitations – as we will see, the healthcare problems and economic constraints vary dramatically throughout the world. With this beginning, we profile new technologies emerging from biotechnology and bioengineering

(1) What are the major health problems worldwide? Global mortality data show a significant gap in health status between developed and developing countries. Leading causes of death in the developed world include cancer, ischemic heart disease, and stroke. In the developing world, infectious diseases like tuberculosis and malaria are far more prevalent owing to widespread poverty, poor infrastructure, and a lack of healthcare resources. A child born today in one of the least developed countries is more than 1000 times more likely to die of measles, an easily preventable and curable disease, than one born in an industrialized country. Worldwide, more than 31 million adults and 2.0 million children are living with HIV/AIDS, most in developing countries. Over the next decade, noncommunicable diseases such as diabetes and heart disease are expected to overtake infectious diseases and malnutrition as leading causes of death in developing countries. The fraction of the global burden of disease linked to lifestyle and behavior choices, currently 20–25%, is expected to increase throughout the world – for example, by 2020 tobacco is expected to kill more people than any single disease, even HIV/AIDS [2]. Understanding how health needs differ throughout the world and how these needs are projected to change in the coming years is the first

Emerging medical technologies

UN Millenium Development Goals Some 80% of the world’s population live in developing countries. In 2000, 189 countries committed to a broad set of goals to meet the needs of the world’s poorest citizens. The goals include the following. Eradicate extreme poverty and hunger r Halve the proportion of people whose income is less than one dollar a day by 2015. r Halve the proportion of people who suffer from hunger by 2015. Achieve universal primary education r Eliminate gender disparity in primary and secondary education in all levels of education by 2015. Reduce child mortality r Reduce the under-five mortality rate by two thirds by 2015. Improve maternal health r Reduce the maternal mortality ratio by 75% by 2015. Combat HIV/AIDS, malaria and other diseases r Halt and begin to reverse the spread of HIV/AIDS by 2015. r Halt and begin to reverse the incidence of malaria and other major diseases by 2015. Ensure environmental sustainability r Halve the proportion of people without sustainable access to safe drinking water and sanitation by 2015. Develop a global partnership for development The Millenium Country Profiles ( provide a source of data to compare economic and health status of countries and to monitor progress toward these goals [1].


Table 1.1. Average health care expenditures per capita of selected WHO nations [3]. Country

Avg. Health Care Expenditure per capita, 2001 (US$)

















United States


step to enable the development of new technologies to address these needs.

(2) Who pays to solve problems in healthcare? Despite recent advances, many medical technologies are available only to a small segment of the world’s population. As a result, standards of medical care differ radically between the developed and developing world. Average annual healthcare expenditures in high income countries are more than $1800 per person, compared to only $16 per person in the world’s least developed countries (Table 1.1). Even in high income countries, the cost of new medical technologies is of great concern. Over the past two decades, healthcare spending has risen dramatically in the United States and throughout the industrialized world, and this rise is expected to continue through the next decade. In the USA, healthcare costs now account for one seventh of the nation’s expenditures. The increasing use of new, expensive technologies, an aging population, and increased administrative costs all contribute to the overall rise in healthcare spending. As we will see later, increasing health expenditures does not always improve health status. As health spending grows beyond a minimum value, there is a decreasing rate of return on investment, with fewer years of life gained per dollar invested [4]. In order to achieve the promise of new technologies worldwide,


Biomedical Engineering for Global Health

Major areas of bioengineering Tissue engineering and regenerative medicine The use of engineering design principles to regenerate natural tissues and create new tissues using biological cells and three dimensional scaffolds of biomaterials. Molecular and cellular engineering Engineering approaches to modify properties of molecules and cells to solve biotechnological and medical problems. Computational bioengineering Use of computational tools to analyze large biological data sets such as in genomics or proteomics; computational models to predict structure and behavior of large biological molecules and to guide design of new drugs. Biomedical imaging Design of imaging systems (e.g. ultrasound), image analysis tools, and contrast agents to record anatomic structure or physiologic function. Biomaterials The engineering design of materials compatible with biological organisms that can be used to make implants, prostheses, and surgical instruments that do not provoke immune rejection. Drug delivery Design of materials and systems to achieve controlled release of drugs in physiologic systems. Biomechanics The study of mechanical forces in living systems and the use of engineering design to create prosthetic devices and tools for rehabilitation. Biosensors Engineering design of systems to identify and quantify biological substances. Advances in microelectronics have aided in developing miniature, implantable biosensors. Biosystems engineering Modeling complex, interacting networks of biological systems within cells and organisms to understand physiology and disease and suggest therapeutic strategies to modify behavior.

our society must develop and evaluate technologies in a cost-conscious manner.

(3) How can bioengineering solve global health problems? Technology development begins with scientific knowledge; in health issues this often means an understanding of a disease and its effects on the body. Bioengineers build on this scientific knowledge to create new technologies that solve healthcare problems. Magnetic

resonance imaging, radiation therapy, and vaccines are all examples of health-related technologies that have become widespread within the past century. The heart– lung bypass machine, pacemakers and other technologies have revolutionized the treatment of heart disease, reducing cardiovascular mortality by half over the past 50 years. In this book, we will consider how new technologies can be used to diagnose, treat, and ultimately prevent the three leading causes of death throughout the world: infectious disease, cancer, and heart disease.

Emerging medical technologies


As we will see later, the development of new healthcare technologies must take into account the societal and economic context in which they will be used and their

Answers to these four questions are complex and interrelated. We begin our journey to understand how bioengineering can be used to improve world health

potential status as a priority or a luxury at a given time. For example, development of a totally implantable artificial heart may provide a solution to the problem of endstage heart failure in developed countries, but owing to differences in infrastructure and resources is unlikely to be a practical solution in many developing countries. To help illustrate these challenges, throughout this book, we will profile the experiences of several under-

by examining a case study of the development of a new technology – the use of high dose chemotherapy and bone marrow transplant to treat advanced breast cancer. This case study illustrates the difficult personal and social issues that can arise as new technologies are developed and tested, and will introduce many of the issues that we will examine in more detail throughout the text. We conclude our case study with a look at how

graduate students who carried out internships in subSaharan Africa as a part of a course in Bioengineering and World Health. Their experiences highlight both the opportunities and challenges of developing new technologies to improve world health.

the process of healthcare technology assessment can be systematically used to address these complex and sensitive issues in a scientifically sound manner.

(4) How do new technologies move from the laboratory bench to the patient’s bedside? New medical technologies developed in research laboratories must be subjected to a rigorous testing procedure to ensure that they are both safe and effective. In many cases, this involves carrying out experiments with human subjects. How can we ensure that these experiments are carried out in an ethical way? How can we balance the desire to bring promising new treatments to patients who need them as soon as possible against the risk of harming patients by allowing them access to therapies that haven’t been sufficiently tested? As healthcare consumers we are often faced with conflicting media reports of the safety of new medical technologies. In order to make choices about our own healthcare, it is necessary to understand how medical research is funded and how new drugs and medical devices are regulated.

Learn more about breast cancer Breast Cancer Facts and Figures 2005–2006. (Atlanta, GA: American Cancer Society, Inc.; 2005). [5] CAFF2005BrF.pdf

Case study: breast cancer and bone marrow transplant Breast cancer is both a devastating and a common disease. If you are female and live in the United States, you have a one-in-eight (12.5%) chance of developing breast cancer sometime in your life [5]. When detected early, there are many effective treatments for breast cancer. However, few effective treatments exist for the disease in its later stages. Less than 20% of women are alive five years after the detection of Stage IV metastatic breast cancer, the most advanced form of the disease. In the 1980s a promising new therapy was developed for women with metastatic breast cancer: high dose chemotherapy followed by bone marrow transplant (HDCT+BMT). Small, early clinical trials of this technique were very promising. The effectiveness of a new cancer treatment is initially measured by the fraction of patients who experience a complete or total response following treatment. In the 1980s, a number of small studies showed a substantial increase in the number of patients with metastatic breast cancer who responded to this new therapy compared to historical experience for patients treated with standard chemotherapy. Although these results were exciting, they were viewed with caution until the patients could be followed for a longer period of time. Many patients who initially respond to therapy may relapse; thus long term


Biomedical Engineering for Global Health

Beyond Traditional Borders: Reports from Student Interns Kim Bennett accompanied Dr. Ellie Click across Malawi conducting intensive training at hospitals as part of a pilot project for the use of bloodspot PCR for infant HIV diagnosis. Dave Dallas and Tessa Elliott assisted in the design and implementation of World Food Program food distribution system at a pediatric AIDS clinic in Mbabane. Lindsay Zwiener and Rachel Solnick pilot-tested software that generates pictorial medication guides, which were developed as their Bioengineering & World Health course projects. They assessed whether these guides help caregivers in Botswana in the proper dosing and timing of anti-retroviral (ARV) medications, promoting adherence to ARV therapy. Christina Lagos and Sophie Kim rolled out their Bioengineering & World Health course project in the SOS Village in Maseru. The project was an after-school activities club to promote interest in science and health education with a focus on HIV/AIDS. They also implemented a Reach Out and Read program at a pediatric AIDS clinic. The course in Bioengineering & World Health was developed and offered at The University of Texas at Austin and at Rice University. Through a new initiative called Beyond Traditional Borders, made possible by a grant to Rice University from the Howard Hughes Medical Institute through the Undergraduate Science Education Program, students at Rice University can travel to Africa for a summer and implement the projects they developed as part of this course. The inaugural class of interns kept a blog describing their experiences. Throughout the book, we include excerpts from the blog to provide a student’s view of how bioengineering can improve world health. You can find more student blogs at:

Emerging medical technologies


Departure: June 8th, 2007 Christina Lesotho Coming from a close family, I have been doing a lot of explaining about my goals and purpose for this trip and doing my best to calm the fears of my family. I know that they simply want me to be safe and are concerned about me while I am gone, and I am used to the ways of overprotective Greek relatives. In the end, I think I have convinced them that this will be the experience of a lifetime and that I have been looking forward to something like this since I began college. I was getting ready to record something in my personal journal last night and found that the last sentence I wrote the last time I made an entry had to do with Africa. From my last weeks in Washington, D.C., working on health policy in Africa, I expressed a desire to go and experience the challenges and situations first hand. “I want to go to Africa . . . why not me?”, that is what I had written as I wondered why it always seemed so far-fetched or impossible that I would one day be able to visit. And now it’s quickly approaching, and I feel so fortunate and excited for this opportunity. I am prepared for some of the best and worst emotions I have ever experienced and am ready to fully immerse myself in the work I am about to do in Lesotho. I feel almost guilty for having somehow cheated during this pre-departure period . . . I have been looking at tons of Google images of Maseru, Lesotho, and the surrounding area, and I feel like I have some sort of unfair advantage as I travel. When I was younger and did not use or have access to the Internet as much, traveling to a new place was always so much more of a mystery and I always envisioned my destination so differently than it turned out to be. I know that a bunch of Google images and travel sites will not do Lesotho justice, but I still feel like I have done away with at least a bit of the mystery of travel. Maybe I won’t do that next time. I am looking forward to spending the next few days in Johannesburg with a family-friend who grew up there. I will be there until the 12th when I will be meeting up with Sophie at the airport to head to Maseru. It will be nice to leave the hot and humid start of summer here in Florida and find the cold beginnings of winter in southern Africa!

survival rates are often used as a better metric to determine the effectiveness of a new cancer therapy. The

was little clinical evidence to show that it was superior to standard therapy [8]. The story of what happened

three year survival rate measures the number of patients still alive three years after beginning cancer therapy. In the early 1990s, a small study indicated that women with high risk breast cancer treated with HDCT+BMT had a 72% three year survival rate, dramatically higher than the historical experience for women treated with standard dose chemotherapy, which was only 38–52% [6]. These studies offered new hope to women who faced

as this technology was developed and tested illustrates how political pressures can overwhelm science, leading to substantially increased medical costs and dramatically reduced quality of life for patients.

high risk or metastatic breast cancer. HDCT+BMT is a grueling treatment that has been described by Dr. Jerome Groopman as “an experience beyond our ordinary imaginings – the ordeal of chemotherapy taken to a near-lethal extreme [7].” In desperation, more than 41 000 American women with advanced breast cancer endured HDCT+BMT in the 1990s, even though there

Breast cancer in the USA After skin cancer, breast cancer is the most common cancer among women, and accounts for almost one of every three cancers diagnosed in women in the United States [5]. In 2005, more than 40,000 American women are expected to die of breast cancer; only lung cancer causes more cancer deaths in women. An estimated 211,240 new cases of breast cancer occurred in the USA in 2005, and there are over 2.3 million women living in the USA who have been diagnosed with breast cancer.


Biomedical Engineering for Global Health

Figure 1.2. Female breast cancer incidence rates by race and ethnicity in the United States as reported by SEER [9]. The rates are age adjusted to the 2000 USA standard population.

Figure 1.3. Female breast cancer death rates by race and ethnicity in the United States as reported by SEER [10]. The rates are adjusted to the 2000 USA standard population.

Emerging medical technologies


Female breast cancer incidence rates have risen in the USA from 1973 to 1998, as reported by the NCI Surveillance, Epidemiology and End Results (SEER) Program (Figure 1.2). Incidence rates have increased owing to a combination of changes in reproductive patterns (delayed childbearing, having fewer children) and better early detection with mammography. Female breast cancer death rates in the USA during the same period have decreased (Figure 1.3), primarily owing to better early detection of more treatable cancers and to improvements in breast cancer treatments. Figure 1.4 shows an illustration of the female breast. After childbirth, milk is produced in glandular tissue in the breast, leading to milk ducts [11]. This glandular tissue is where most breast cancers develop. When cancer cells are confined to these ducts, and have not spread


Thymus Gland


Lymph Nodes

to surrounding fatty tissue, the disease is called Stage 0, and is completely curable with surgical excision. Lesions which have spread to the surrounding fatty tissue but

Lymphatic Vessels

Chest wall

Pectoralis muscles


Figure 1.5. Lymphatic system. Source: SEER Training Modules, Lymphoma. US National Cancer Institute. 2009.

Nipple surface Areola duct

are less than 2 cm in diameter are referred to as Stage I

Fatty tissue Skin

Figure 1.4. The human female breast. Source: SEER Training Modules, Breast Cancer. US National Cancer Institute. 2009.

lesions, and also have excellent prognosis, with a 100% five year survival rate [12]. A series of lymphatic vessels, leading to lymph nodes under the armpit (axillary lymph nodes), drain breast tissue (Figure 1.5) [11]. Breast cancer cells can migrate from the initial lesion and enter these lymphatic vessels, providing a way for breast cancer cells to spread to other distant organ sites (metastasize). If the cancer has spread to one–three lymph nodes close to the breast but not to distant sites, it is referred to as a Stage II lesion, and the five year survival rate is in


Biomedical Engineering for Global Health

Table 1.2. Breast cancer staging [12]. Stage


5 yr survival

Stage 0

Cancer cells are located within a duct and have not invaded the surrounding fatty breast tissue


Stage I

The tumor is 2 cm or less in diameter and has not spread to lymph nodes or distant sites.


Stage II

The cancer has spread to 1–3 lymph nodes close to the breast but not to distant sites


Stage III (High risk)

The cancer has spread to 4–9 lymph nodes close to the breast but not to distant sites


Stage IV Cancer has spread to distant organs (Metastatic) such as bone, liver or lung or to lymph nodes far from the breast.


Reprinted by permission of the American Cancer Society, Inc. from All rights reserved.

the range 81–92%. Stage III breast cancers involve more than four nodes, and because the five year survival rates are so low (54–67%) are referred to as “high-risk breast cancers.” In metastatic breast cancer (Stage IV), the disease has spread from the lymphatics to other organ sites far from the breast, such as the brain. The five year survival rate for metastatic breast cancer is only 20%. The stages of breast cancer and the prognosis for each stage are summarized in Table 1.2 [12].

Treatments for breast cancer There are many treatments for breast cancer. Treatment for most early cancers involves some form of surgery to remove the cancer cells. If the lesion is small, only a portion of tissue may be removed (lumpectomy), or the entire breast may be removed (mastectomy). Larger tumors may be treated using chemotherapy. In some cases, chemotherapy may be used to shrink larger tumors so that they can be removed surgically; in others it may be used following surgery to reduce risk of recurrence. In chemotherapy, drugs which are toxic to cancer cells are given intravenously or by mouth. These drugs travel through the bloodstream, reaching cancer cells throughout the body. Chemotherapeutic drugs interfere with ability of cells to divide; many cancer

cells cannot repair damage caused by chemotherapy drugs so they die. Rapidly dividing normal cells may also be affected by chemotherapy drugs, but they can repair this damage. Because chemotherapy drugs affect rapidly dividing normal cells, they give rise to many undesirable side effects. The cells which line the gastrointestinal tract divide rapidly; thus chemotherapy can lead to nausea, vomiting, mouth sores and loss of appetite. Cells in the hair follicles divide rapidly and chemotherapy can lead to hair loss. Rapidly dividing cells in the bone marrow which produce oxygen carrying red blood cells, infection fighting white blood cells, and platelets important in blood clotting are also affected by chemotherapy drugs. Chemotherapy patients are thus at high risk for infection, bleeding and fatigue. While these side effects are temporary, chemotherapy can also produce permanent side effects such as premature menopause and infertility.

High dose chemotherapy Because chemotherapy can damage both cancer cells and rapidly dividing, but crucial, normal cells, cancer treatment must strike a balance between completely destroying all cancer cells while causing minimal damage to normal cells. In the 1980s a number of dose comparison studies of chemotherapy to treat metastatic breast cancer showed that a higher dosage of chemotherapy was associated with a higher response rate. Scientists and clinicians hypothesized that metastatic breast cancer could be treated more effectively with higher doses of chemotherapy. Unfortunately, such high doses completely destroy the bone marrow, leaving patients with no way to continue to produce the cells of the blood system and the immune system, which are necessary for life. Our blood consists of four components: plasma, red blood cells, white blood cells, and platelets. Plasma carries nutrients and hormones throughout the body. Red blood cells deliver oxygen throughout the body, while white blood cells are necessary to fight infections. Platelets are necessary for blood clotting following injury. Throughout our lives, our blood cells are continually renewed within the bone marrow. The source

Emerging medical technologies


Figure 1.6. Red and white blood cells are derived from cells of the bone marrow.

of all these cells is the pluripotent hematopoeitic stem cell which can give rise to all the types of blood cells (Figure 1.6). Laboratory experiments in mice show that a single stem cell can yield the half-trillion blood cells of an entire mouse. Clinicians theorized that if the bone marrow was completely destroyed in high dose chemotherapy, a bone marrow transplant could be done to restore these hematopoeitic stem cells; in fact such bone marrow transplants had proven very successful in the treatment of cancers of the bone marrow.

Learn more about bone marrow transplants In his article, “Bone marrow transplant: a healing hell,” Dr. Jerome Groopman describes the experience of two patients who undergo bone marrow transplants [7].

(PBSCs) which are found in the blood. In a transplant, these stem cells are isolated from the blood in a process known as apheresis. The patient is given medication to increase the number of stem cells in the bloodstream. Next, blood is removed from the body through a central venous catheter and passes through a machine that removes the stem cells (Figure 1.7). The blood is then returned to the patient and the collected stem cells are stored for future transplantation. The entire process takes 10–12 hours, and yields enough stem cells to fill one syringe. The initial attempts to transplant bone marrow took place in Cooperstown, NY during the 1950s [7]. The effects of the atom bomb used at the end of World War II sparked a tremendous interest in identifying ways

Bone marrow transplants

to restore bone marrow. One reason that the bomb’s radiation was so deadly was because it destroyed the bone marrow cells of its victims, leading to hemorrhage (uncontrolled bleeding) and the inability to fight off infection. At the time physicians could successfully transfuse oxygen carrying red blood cells from compatible donor to needy recipient. However, bone marrow cells could not be transfused.

Stem cells are found in high concentration in the bone marrow, and can be harvested for transplantation in a painful procedure. More recently, stem cells transplants have been carried out using peripheral blood stem cells

Invariably, the recipient’s body identified them as foreign invaders and destroyed them [7]. One researcher who was especially interested in the bone marrow transplant problem was Don Thomas.

J. Groopman. Bone marrow transplant: a healing hell. The New Yorker (19 October 1998), pp. 34–39.


Biomedical Engineering for Global Health

Apheresis technology Most stem cells are found in the bone marrow, but some, called peripheral blood stem cells, can be found in the blood. It is typically much more difficult to harvest bone marrow than cells in peripheral blood – harvesting bone marrow requires hospitalization and general anesthesia. Typically, the concentration of stem cells in the peripheral blood is very low, so patients are given growth factors to increase the concentration of peripheral blood stem cells for several days prior to harvesting stem cells. During apheresis, blood is removed from a large vein in the arm and sent to a machine which contains a centrifuge to separate white blood cells. Anticoagulants must be added to the blood to prevent it from clotting. The centrifuge spins the entering blood, and the resulting centrifugal force separates the various components of blood – plasma, red blood cells and white blood cells – based on differences in their density. The red blood cells are pushed to the outside of the centrifuge, while plasma remains near the center of the rotor. A layer of white blood cells called buffy coat separates the plasma and red blood cells. This layer contains the peripheral blood stem cells and is separated. The remaining blood is returned through a tube to the patient’s other arm. A successful transplant requires collection of a large number of peripheral blood stem cells – approximately five million stem cells per kilogram of body weight are required. Thus, we must quantify the number of peripheral blood stem cells harvested during apheresis to determine whether a sufficient number have been collected. When viewed through a standard microscope the stem cells can’t be differentiated from other white blood cells. However, stem cells express a protein called CD34 on their membrane. The fraction of CD34 positive cells can be quantified by labeling the cells with a fluorescent dye linked to a molecule that binds to CD34 and using a special machine called a flow cytometer to count the number of CD34 positive cells. Over 20 liters of blood must be processed (the entire blood volume must be treated four times) to collect sufficient cells for later transplant, and apheresis is typically performed over several days. These cells are then treated with cryopreservatives and frozen to be injected into the patient following the high dose chemotherapy procedure [13, 14].

Thomas treated patients with cancer of the bone marrow (leukemia) with chemotherapy. He believed that

eign invaders and the patient’s own cells. The histocompatibility markers explained the failure of previous

providing new, healthy bone marrow cells was essential to curing leukemia. He tested various transplant techniques in dogs initially, and then in patients with late stage leukemia. In early trials, every patient who underwent transplantation died. “Things were pretty grim”, Thomas later remarked [7]. After four years of unsuc-

transplant attempts and held the key to future success. When not properly matched, the patient’s immune system would reject transplanted cells. Proper matching of histocompatibility markers between donor and recipient led to successful results in dogs. With this advance, Thomas resumed human trials, which led to successful

cessful transplantations attempts, he stopped human trials. Eight years later, Thomas identified protein markers on the surface of white blood cells [7]. These his-

treatment for leukemia. Thomas (Figure 1.8) received the Nobel Prize in 1990 for his important work in this area. Today bone marrow transplantation is a successful

tocompatibility markers are unique to each individual and are found on the surface of nearly every cell in the body, but are particularly numerous on the surface of white blood cells. Histocompatibility markers enable a patient’s immune system to differentiate between for-

treatment for leukemia. In the past 40 years the five year survival rate for leukemia has more than tripled, from 14% in 1960–63 to 49% in 1995–2002 [15]. However, it is still a gruelingly difficult treatment. Dr. Jerome Groopman describes the experiences of two patients

Emerging medical technologies


diagnosed with leukemia. She received a bone marrow transplant, and recounts her experience in Groopman’s article. “It was a complete nightmare. For days, I’d be on all fours and just retch and retch. I looked like a lobster, and thought I had bugs crawling on me. I’d hit myself and scream. I was in that sterile bubble, and forgot what skin against skin felt like. That was lost. I just wanted to hold on to my mom or dad, like a two-yearold, and I couldn’t. I had terrible diarrhea, a blistering rash all over my body, and jaundice. I was the color of an egg yolk [7].”

A new technology for advanced breast cancer, HDCT+BMT With the success of bone marrow transplant for leukemia, clinicians hypothesized that extremely high dose chemotherapy could be used to treat metastatic breast cancer if followed by a bone marrow transplant. In this case, the patient’s own stem cells could be harvested prior to the chemotherapy and then reinfused Figure 1.7. An apheresis machine. Copyright Caridian BCT, Inc. 2009. Used with permission.

following treatment, thus insuring a perfect histocompatibility match. Compared to standard chemotherapy,

who received bone marrow transplants in his article, “Bone marrow transplant: a healing hell [7].” Courtney

this procedure was initially very expensive (>$140,000) and initial trials had very high treatment associated mortality (death) rates, in the range 7–22% [16, 17]. Despite the extreme expense and side effects, the combination

Stevens was a high school sophomore when she was

Figure 1.8. Don Thomas and his wife and partner in research, Dottie, with childhood leukemia survivors. Used with permission c Jim Linna. from 


Biomedical Engineering for Global Health

of HDCT+BMT offered some of the only promise for the treatment of metastatic breast cancer. An early study showed that the three year survival rates of women with high risk breast cancer treated with HDCT+BMT were 40% higher than those of women who had not participated in the trial and had received standard chemotherapy [8]. While this study offered hope for the new treatment, it was criticized for several reasons. It was a small study, involving only 85 patients, and did not randomly assign women to receive either the new therapy or the standard therapy. It also only included women whose disease initially responded to standard chemotherapy and who therefore might be expected to do better than those whose disease was not responsive to standard treatment.

Learn more about HDCT+BMT Drs. Michelle Mello and Troyen Brennan provide a more complete account of the early controversy surrounding the use of HDCT+BMT to treat breast cancer [8]. M. M. Mello and T. A. Brennan. The controversy over high-dose chemotherapy with autologous bone marrow transplant for breast cancer. Health Affairs, 20:5 (2001), 101–17.

In order to gain more evidence, several larger clinical trials were initiated in which women with advanced breast cancer were to be randomly selected to receive either standard chemotherapy or HDCT+BMT. Clinicians planned to compare the percentage of patients who were still alive (survival rates) three and five years following therapy in both arms of the trial as well as the percentage of patients whose cancer had not recurred (disease-free survival rates). Such randomized clinical trials are considered to be the most important kind of clinical evidence to indicate whether a new therapy is better, the same, or worse than a standard therapy. Typically, in the absence of such evidence, a therapy is considered to be experimental and most insurance companies in the USA will not pay for it. Because there are so few effective treatments available for advanced breast cancer however,

there was a strong public demand for HDCT+BMT, even in the absence of good clinical evidence to indicate that it worked.

Public reaction to new hope In 1991, the television show 60 Minutes aired a piece decrying the company Aetna’s decision to deny insurance coverage for HDCT+BMT to treat breast cancer [8]. At the same time, Nelene Fox, a 38 year old mother of three who was diagnosed with advanced breast cancer, sued her insurance company [8]. The company, HealthNet, refused to pay for HDCT+BMT for Fox, even though it had recently paid for a relative of its CEO to receive the same treatment. Mrs. Fox and her family sued HealthNet for failure to provide coverage. In the meantime the family raised more than $210 000 so she could receive HDCT+BMT. Mrs. Fox died of breast cancer before a verdict was reached; her family argued that the delay in receiving the treatment contributed to her death. The family was awarded $89M, then the largest jury verdict ever against an HMO. The case received widespread publicity, and in 1993 the Massachusetts legislature mandated that insurers provide coverage for HDCT+BMT for advanced breast cancer. In 1994, insurers approved coverage for 77% of breast cancer patient requests for HDCT+BMT as part of clinical trial participation [8]. However, approval was highly arbitrary, even for similar patients covered by the same insurer. Nine of 12 large insurers surveyed indicated that the threat of litigation was a major factor in their decision to provide coverage. In 1995, the results of a small, short randomized trial of 90 patients in South Africa was reported by the lead physician, Dr. Werner Bezwoda [20]. Dr. Bezwoda’s study showed that, on average, women who received HDCT+BMT for metastatic breast cancer survived twice as long without a relapse than women who received standard chemotherapy. By this time, more than 80% of American physicians believed that women with metastatic breast cancer should be treated with HDCT+BMT, and these results seemingly supported that conclusion [8]. During the 1990s, more than 41,000 patients underwent HDCT+BMT for breast cancer despite a paucity of clinical evidence regarding

Emerging medical technologies


Breast cancer in developing countries

Learn more about insurance coverage for HDCT+BMT

More than 1.2 million people worldwide were diagnosed with breast cancer in 2005. Women in developed countries have access to imaging technologies such as mammography and ultrasound

The Aetna insurance company will only pay for HDCT+BMT as part of a controlled clinical trial sponsored by the Food and Drug Administration or the National Cancer Institute. This article explains

to aid in early detection and to advances in hormonal treatments and chemotherapy. However, women in developing countries frequently do not have access to these lifesaving technologies. Maria Saloniki is a 60 year old mother of ten

their rationale [19]. Breast cancer: high dose chemotherapy with autologous sterm cell support. Clinical Policy Bulletins (Aetna Inc, 7 October 2005).

living in the United Republic of Tanzania. This image of her (below) is used courtesy of WHO/ Chris de Bode. When she was 57 she experienced fever, a swollen armpit and pain. Over three years, she visited local healers, various clinic doctors, and even traveled to Nairobi, Kenya to seek treatment. She was prescribed herbal ointments, antibiotics,

effectiveness. In fact, it was so difficult to recruit patients to randomized phase III clinical trials (because

and told that nothing could be done for her condition. Finally, three years after her initial symptoms she traveled to Dar es Salaam, where a biopsy showed that she had breast cancer and she began chemotherapy. Her husband has had to borrow a large sum to finance her care, and can’t afford both the cost of the treatment and bus fare to come and visit her [33].

women were afraid they would be randomly selected to receive the standard therapy) that the trials took more than twice as long to complete than planned. In 1999, at the meeting of the American Society of Clinical Oncology, the results of five randomized clinical trials were reported. Sadly and surprisingly, four of the studies showed no survival benefit with HDCT+BMT; some showed it took a little longer for cancer to return. Figures 1.9a and 1.9b compare the survival and disease free survival rates over time in women receiving either HDCT+BMT or standard therapy in one of the trials; no meaningful differences were noted in either case. Only one South African study, again from Dr. Bezwoda, showed a survival benefit [22]. In his study, women with high-risk breast cancer had an 83% chance of five year survival if they received HDCT+BMT, compared to only a 65% chance of five year survival with standard chemotherapy. The average disease free survival time was 100 months for women receiving HDCT+BMT, versus only 47.5 months average disease free survival for those receiving standard chemotherapy. The poor results of the four negative HDCT+BMT trials were widely reported in the media. Public reaction was again strong. Prior to the negative trial results, in 1996–98, Anthem Insurance saw the number of women requesting HDCT+BMT for breast cancer increase [25]. In 1999, prior to the trial results, the company expanded indications for which


Biomedical Engineering for Global Health


Scientific misconduct Scientists could not understand why one trial showed improved survival with HDCT+BMT, while four other trials showed no benefit. A team of scientists was sent to


audit the results of the South African trial (Figure 1.10). Unfortunately, the audit team could not find records for many of the patients supposedly enrolled in the study. They found that the study showed little evidence of randomization, and that many patients whose records could be found did not meet the eligibility criteria for the trial [26]. They also found that the trial had not been properly approved by the Institutional Review Board at Dr. Bezwoda’s university, which is required to approve all research involving human subjects in advance. The university conducted a formal ethics inquiry, and Dr. Bezwoda admitted to a “serious breach of scientific honesty and integrity [27].” The university fired Dr. Bezwoda, and many of his publications were formally retracted from the journal in which they had been published.

Figure 1.9. Disease free survival (a) and survival (b) rates over time in women receiving either HDCT+BMT or standard therapy from a c 2003 Massachusetts randomized clinical trial [21]. Copyright  Medical Society. All rights reserved.

Learn more about public reaction Denise Grady covered the announcement of these results for the New York Times [23]. D. Grady. Doubts raised on a breast cancer procedure. New York Times (16 April 1999). Joanne Silberner covered the announcement of these results for National Public Radio [24]. J. Silberner. Breast cancer. Radio program, National Public Radio (16 April 1999). story.php?storyId=1049404.

they would approve HDCT+BMT. After the trial results were reported in 1999, they received only four requests for such coverage, despite the expanded coverage. Most insurance companies now cover HDCT+BMT for breast cancer only as part of an FDA or NCI sponsored clinical trial.

Where are we now? Scientists continued to follow the patients enrolled in the five randomized clinical trials originally reported in 1999 (Table 1.3). Even with longer follow up, it appears that there is no survival benefit to HDCT+BMT at either three years or five years following treatment as compared to standard chemotherapy. There is a small but significant increase in disease free survival at three years with HDCT+BMT, but this advantage disappears at five years. Serious side effects are more common with HDCT+BMT compared to standard therapy, but most are reversible. Patients report that quality of life is lower at six months following treatment with HDCT+BMT, but similar to that of standard chemotherapy one year following treatment. The costs of HDCT+BMT have been reduced to about $60,000, which is still nearly two times that of standard chemotherapy. Most physicians and insurance companies now agree that HDCT+BMT should not be used to treat high risk breast cancer outside of a randomized clinical trial. Research in this area continues, to identify if longer follow up (7–10 years) will show advantages of high dose therapy, or to determine if there are sub-groups of

Emerging medical technologies


Table 1.3. Results of five randomized clinical trials of HDCT+BMT for breast cancer [21, 28–30]. Study

# Randomized patients

% survival

Disease free survival

Stadtmauer Metastatic


32% 3 year BMT 38% 3 year control

9.6 months BMT 9.0 months control

Lotz Metastatic


29.8% 5 year BMT 18.5% 5 year control

9% disease free at 5 yrs BMT 9% disease free at 5 yrs control

Peters High Risk


79% 3 year BMT 79% 3 year control

71% disease free at 3 yrs BMT 64% disease free at 3 yrs control

Rodenhuis High Risk


75% 5 year BMT 73% 5 year control

65% disease free at 5 yrs BMT 59% disease free at 5 yrs control p = 0.09∗

Tallman High Risk


58% 6 year BMT 62% 6 year control

49% disease free at 6 yrs BMT 47% disease free at 6 yrs control

Figure 1.10. The results of Dr. Bezwoda’s controversial trial [26].

women who benefit from high dose therapy (for example those whose tumors are negative for certain genetic markers or who have 10 or more axillary lymph nodes which show cancer cells). New technologies to completely rid the transplanted stem cells of any rogue cancer cells may also reduce recurrence rates in women treated with HDCT+BMT. However, all of these theories must be subject to rigorous testing if they ever are

media, the courtroom or the laboratory? In an age where high-technology treatments are one of the most powerful drivers of healthcare costs, these are crucial questions.

Healthcare technology assessment

Lessons learned

Professors Frazier and Mosteller, experts in health policy and management, have stated, “If we are to have good medical care, we need to know what works, and this cannot be known without systematic technology assessment. The intuitions of physicians and the guesses

The example of HDCT+BMT to treat breast cancer illustrates the dangers of allowing political pressures to overwhelm scientific evidence. What is the proper forum to resolve such controversies? Should it be the

of biologists are not adequate guides to the best treatments [31].” How then do we assess new technologies objectively, avoiding political pressures that can lead us to waste precious healthcare resources and subject

to become methods of standard treatment.


Biomedical Engineering for Global Health

thousands of patients to punishing, but ineffective, treatments?

Healthcare technology assessment The systematic process of evaluating the safety, short term and long term efficacy, acceptability and cost effectiveness of a new medical technology.

Answering these questions is increasingly important in a world where early studies of new medical advances can receive substantial publicity in the popular press before randomized clinical trials are completed. A recent study published in the Journal of the American Medical Association compared conclusions presented in highly cited articles in major general clinical journals to those of subsequent studies with larger sample size or better controlled design. Results showed that nearly 1/3 of highly cited studies were later contradicted and that this was most likely for nonrandomized studies [32]. As we examine these important issues in this book, we will build a toolkit to help us answer politically sensitive questions about how to use limited resources in a deliberate and unbiased manner. Technology assessment will be an important part of our toolkit, and it is the subject of Chapter 2.

Bioengineering and Global Health Project Project overview Design a new technology to solve a health problem, present a mock prototype of the new technology to a design review committee, and design a clinical trial to test the new technology. Throughout this text, you will use the engineering design method to design a new solution to an important health problem. You will identify an important health problem, and carry out research to understand the scope of the problem and limitations of current health technologies. You will follow the engineering method to design a new solution which meets the constraints you identify. You will create a physical prototype of your design and will present it to the class as part of a design review exercise.

Homework 1. Advanced breast cancer has a high mortality. Initial clinical trials indicated that high dose chemotherapy followed by a bone marrow transplant could reduce the mortality rate by as much as 40%. a. Why did physicians and scientists believe that higher doses of chemotherapy would be more effective than standard therapy for advanced breast cancer? b. Why is it necessary to give patients a bone marrow transplant following high dose chemotherapy? What will happen if they do not receive a bone marrow transplant? c. In the context of this example, discuss how political pressures overwhelmed scientific evidence. How could this be avoided in the future? d. Find a news report describing a new health technology published in the last year. In your opinion, does this news report provide balanced discussion of the potential promise and the potential limitations of this technology? 2. The Pew Global Attitudes Project is a worldwide survey of public opinion. In 2002, more than 38,000 people in 44 countries were asked to assess the quality of their own lives, their level of optimism about their lives in the next five years, and to rank problems faced by themselves and their countries. In this exercise, you are asked to review the results of this survey and to prepare several graphs summarizing the results. Pew World Attitudes Website: http://people-press. org/reports/display.php3?ReportID=165 Pew World Attitudes Report: You will examine results in countries profiled in Unit 2: the United States, Canada, China, India and Angola. For parts a–e, please construct graphs, for part f provide a discussion which supports your findings. a. What fraction of people surveyed in each country expressed satisfaction with their own lives? b. What fraction of people surveyed in each country report that they are unable to afford food?

Emerging medical technologies c. What fraction of people in each country cite the following as a very big problem in their country? Poor drinking water Crime AIDS and disease d. What fraction of people in each country believe that the following is the greatest danger facing the world today? Nuclear weapons AIDS and other infectious diseases e. What fraction of people surveyed in each country are optimistic that their lives will improve in the next five years? f. Compare general agreement on questions 4 and 5 throughout countries in Africa and Europe.




References [1] Health in the Millenium Development Goals: Millenium Development Goals, Targets and Indicators Related to Health. World Health Organization; 2004. [2] Glossary of Globalization, Trade and Health Terms: Health Transition. World Health Organization; 2007. [3] Beaglehole R, Irwin A, Prentice T. The World Health Report 2004: Changing History. Geneva, Switzerland: The World Health Organization; 2004. [4] Coulter SL, Cecil B. Assessing the Value of Health Care in Tennessee. Chattanooga: Blue Cross Blue Shield of Tennessee; 2003. [5] Breast Cancer Facts and Figures 2005–2006. Atlanta, GA: American Cancer Society, Inc.; 2005. http://www. [6] Peters WP, Ross M, Vredenburgh JJ, Meisenberg B, Marks L, Winer E, et al. High-dose chemotherapy and autologous bone marrow support as consolidation after standard-dose adjuvant therapy for high-risk primary breast cancer. Journal Of Clinical Oncology: Official Journal Of The American Society Of Clinical Oncology. 1993 June; 11(6): 1132–43. [7] Groopman J. Bone marrow transplant: a healing hell. The New Yorker. 1998 October 19: 34–9. [8] Mello MM, Brennan TA. The controversy over high-dose chemotherapy with autologous bone marrow transplant for breast cancer. Health Affairs (Project Hope). 2001 Sep–Oct; 20(5): 101–17. [9] Surveillance, Epidemiology, and End Results (SEER) Program ( SEER∗ Stat Database: Incidence – SEER 13 Regs Public-Use, Nov 2005 Sub










(1992–2003) and SEER 9 Regs Public-Use, Nov 2005 Sub (1973–2003), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2006, based on the November 2005 submission. Surveillance, Epidemiology, and End Results (SEER) Program ( SEER∗ Stat Database: Mortality – All COD, Public-Use With State, Total U.S. for Expanded Races/Hispanics (1990–2003) and All COD, Public-Use With State, Total U.S. (1969–2003), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2006. Underlying mortality data provided by NCHS ( Silverthorn DU. Human Physiology: An Integrated Approach. 2nd edn. Upper Saddle River, NJ: Prentice Hall; 2001. American Cancer Society. Detailed Guide: Breast Cancer: How is Breast Cancer Staged? Sep 18 2006 [cited 2007]. Available from: content/CRI 2 4 3X How is breast cancer staged 5.asp Walker F, Roethke SK, Martin G. An overview of the rationale, process, and nursing implications of peripheral blood stem cell transplantation. Cancer Nursing. 1994 Oct 10; 17(2): 141–8. Hansson M, Svensson A, Engervall P. Autologous peripheral blood stem cells: collection and processing. Medical Oncology. 1996 Jun; 13(2): 71–9. Leukemia, Lymphoma, Myeloma, Facts and Statistics 2006–2007. White Plains, NY: Leukemia and Lymphoma Society; 2007. White K. Notebook: bone marrow transplant for breast cancer is questioned on basis of incomplete data. Journal of Women’s Health and Gender-Based Medicine. 1999; 8(5): 577–82. Brockstein BE, Williams SF. High-dose chemotherapy with autologous stem cell rescue for breast cancer: yesterday, today, and tomorrow. Stem Cells. 1996; 14: 79–89. Abegunde D, Beaglehole R, Durivage S, Epping-Jordan J, Mathers C, Shengelia B, et al. Preventing Chronic Diseases: A Vital Investment. Geneva, Switzerland: World Health Organization; 2005. Aetna. Breast cancer: high-dose chemotherapy with autologous stem cell support. Clinical Policy Bulletins 2005 October 5 [available from:]. Bezwoda WR, Seymour L, Dansey RD. High-dose chemotherapy with hematopoietic rescue as primary




[23] [24]



Biomedical Engineering for Global Health treatment for metastatic breast cancer: a randomized trial. Journal Of Clinical Oncology: Official Journal Of The American Society Of Clinical Oncology. 1995 Oct; 13(10): 2483–9. Tallman MS, Gray R, Robert NJ, LeMaistre CF, Osborne CK, Vaughan WP, et al. Conventional adjuvant chemotherapy with or without high-dose chemotherapy and autologous stem-cell transplantation in high-risk breast cancer. The New England Journal Of Medicine. 2003 Jul 3; 349(1): 17–26. Bezwoda WR. Randomised, controlled trial of high dose chemotherapy versus standard dose chemotherapy for high risk, surgically treated, primary breast cancer. Proceedings of the American Society of Clinical Oncology. 1999; 18: 2a. Grady D. Doubts raised on a breast cancer procedure. New York Times. 1999 April 16; Sect. A1. Silberner J. Morning Show: breast cancer. National Public Radio. 1999 April 16. templates/story/story.php?storyId=1049404. Anthem. Insurance payments for bone marrow transplantation in metastatic breast cancer. The New England Journal of Medicine. 2000 April 13; 342(15): 1138–9. New audit uncovers scientific misconduct in 1995 South African study on metastatic breast cancer. American Society of Clinical Oncology. 2001 April 26.

[27] Horton R. After Bezwoda. Lancet. 2000 Mar 18; 355(9208): 942–3. [28] Rodenhuis S, Bontenbal M, Beex LV, Wagstaff J, Richel DJ, Nooij MA, et al. High-dose chemotherapy with hematopoietic stem-cell rescue for high-risk breast cancer. The New England Journal Of Medicine. 2003 Jul 3; 349(1): 7–16. [29] Stadtmauer EA, O’Neill A, Goldstein LJ, Crilley PA, Mangan KF, Ingle JN, et al. Conventional-dose chemotherapy compared with high-dose chemotherapy plus autologous hematopoietic stem-cell transplantation for metastatic breast cancer. Philadelphia Bone Marrow Transplant Group. The New England Journal of Medicine. 2000 Apr 13; 342(15): 1069–76. [30] Antman KH. Randomized trials of high dose chemotherapy for breast cancer. Biochimica Et Biophysica Acta. 2001 Mar 21; 1471(3): M89–98. [31] Frazier HS, Mosteller F. Medicine Worth Paying For: Assessing Medical Innovations. Cambridge, Mass.: Harvard University Press; 1995. [32] Ioannidis JP. Contradicted and initially stronger effects in highly cited clinical research. JAMA: The Journal Of The American Medical Association. 2005 Jul 13; 294(2): 218–28. [33] Face to Face with Chronic Desease: Maria’s Story: Fighting Cancer. World Health Organization, 2005.

2 Bioengineering and technology assessment

In Chapter 1, we examined the development and introduction of a new technology that initially appeared as if it could provide new hope to women with advanced breast cancer. Small clinical trials showed that women with high risk or metastatic breast cancer treated with high dose chemotherapy and bone marrow transplant (HDCT+BMT) had substantially better response rates and survival compared to historical experience with standard chemotherapy. These early promising results were widely publicized, and even though the therapy had serious side effects, it was used to treat thousands of women. Usually, before a new technology is adopted, randomized clinical trials are conducted to compare the performance of the new technology to that of existing technologies. In such randomized clinical trials, patients are randomly selected to receive either the current standard therapy or the new therapy; outcomes such as response rate, survival and side effects are then compared for the two groups of patients. However, because patient demand was so high for HDCT+BMT, randomized clinical trials took much longer to complete than planned. Ultimately, randomized clinical trials showed that HDCT+BMT did not improve survival for most patients. The promising results of early trials were misleading due to a combination of factors, including their small size, selection bias and scientific misconduct.

The case study of HDCT+BMT for advanced breast cancer underscores the need for a systematic method to guide the development and introduction of new technologies. In Chapter 1, we saw how the interplay of desperate patients seeking the best treatment, early media publicity and a scientist who falsified data all combined to slow the progress of medical science. In the end, many patients unnecessarily underwent an expensive and highly toxic therapy. How can we prevent this from happening with future technologies? In this chapter, we will consider the methodology of technology assessment, which provides a systematic set of tools to determine the performance of a new technology and to assess the impact of using the technology both for individual patients and for society as a whole. When used properly, technology assessment can help ensure that new medical technologies are introduced on the basis of sound scientific evidence and not simply on the opinions of physicians and scientists, or the hopes of patients. As a prelude to technology assessment, we consider the steps involved in bringing a new technology from the laboratory bench to the patient’s bedside. Figure 2.1 shows a roadmap of this process. Bioengineers build on the scientific understanding of a disease to design new healthcare technologies. New technologies must be rigorously tested to determine whether they are safe and effective. This testing process can include preclini-


Biomedical Engineering for Global Health

(b) (a)


(d) (e)



Figure 2.1. A roadmap of the healthcare technology development process. Technology assessment spans the entire range of development activities. Parts (b), (e) and (f), source: Wikimedia. Part (c), source: Jupiter Images. Part (d), source: NCI/Lindia Bartlett. Part (g), source: NCI/Michael Anderson.

cal testing in cell or animal models, as well as testing in human subjects. These tests must be carried out in an

The Littenberg method of technology assessment

ethical manner. In addition, an important consideration in the adoption of new technologies is whether they are cost effective. The process of health technology assessment spans all the steps in the healthcare technology

Benjamin Littenberg proposed a model of technology assessment that is particularly useful for new technologies [1]. The Littenberg method asks five questions regarding a new medical technology.

development process, from lab to patient.

Learn more about the Littenberg method The Littenberg method of technology assessment is defined in this article and used to analyze screening tests for hypercholesterolemia [1]. Littenberg, B. Technology assessment in medicine. Academic Medicine, 67(7), 424–428.

r Biologic plausibility: does our current understanding of the biology of the disease in question support the use of the technology? r Technical feasibility: can we safely and reliably deliver the new technology to the target patients? r Clinical Trials: do the results of randomized clinical trials comparing the new technology to current standards of care show a benefit?

Bioengineering and technology assessment


r Patient outcomes: are patients better off for having used the new technology? r Societal outcomes: what are the costs and ethical implications of the technology? It is useful to consider our case study of HDCT+BMT in the context of the Littenberg model to see whether the technology was assessed appropriately at each level and whether that assessment supports the use of the technology. Biologic plausibility: many scientific studies supported the promise of HDCT+BMT. In particular, as the dose of chemotherapeutic agent was increased to treat women with breast cancer, response rates increased. Based on these data, physicians believed patients with advanced breast cancer would benefit from doses of chemotherapy so high that it would destroy bone marrow. Technical feasibility: mortality rates were initially quite high for breast cancer patients treated with HDCT+BMT, despite the advances in leukemia treatments showing that bone marrow transplantation could be safely performed. However, as more women were treated and regimens were refined, mortality rates dropped substantially, improving technical feasibility.

Figure 2.2. Multiple clinical trials have proven mammography to be effective for screening for breast cancer in women. CDC.

Thus initially HDCT+BMT was supported by both biologic plausibility and technological feasibility, the first two criteria of Littenberg’s method.

apy; later trials examined survival. In assessing patient outcomes it is important to consider both short term

Clinical trials: there were many small, clinical trials carried out to assess the effectiveness of HDCT+BMT; however, these trials were not randomized clinical tri-

outcomes (e.g. response rates) and long term outcomes (e.g. survival) as well as quality of life issues. Clearly, patients treated with HDCT+BMT experienced a quality of life that was initially lower because of the side effects

als. As we will see later, a randomized clinical trial is the strongest source of scientific evidence to assess whether a new technology is effective compared to current standards of care (Figure 2.2). The tragedy of HDCT+BMT was the delay of completing randomized clinical trials, due to political and media pressures and scientific misconduct. By the time clinical trial results seriously questioning the benefit of HDCT+BMT compared to stan-

of the treatment. In the end, randomized trials showed that survival rates were not substantially higher than those for standard treatment. Societal outcomes: was society better off for having used HDCT+BMT? The new technology was substantially more expensive than standard chemotherapy while adding no additional survival benefit. If HDCT+BMT had showed clinical benefit compared to

dard chemotherapy were available, many women had already undergone the treatment. Patient outcomes: were women better off in the long term for having been treated with HDCT+BMT? The early clinical trials assessed only response rates to ther-

standard therapy, then society would have to consider the difficult question of whether the increased benefit is worth the additional cost. Thus, HDCT+BMT was not supported by the final three criteria of the Littenberg method, a conclusion


Biomedical Engineering for Global Health

Challenges of anti-retroviral drugs: June 13, 2007 Rachel Botswana As we will see in Chapter 4, patients who are HIV positive are treated with drugs called anti-retrovirals or ARVs. Often, patients must follow complex regimes of drugs. The situation is further complicated for HIV positive children, because drug companies currently do not make ARVs in pediatric doses. So often parents must split pills in half to ensure their children receive the correct dosage. Patients must adhere closely to the schedule of taking their ARVs or HIV can develop resistance to the drugs. This is a major challenge in treating HIV/AIDS today. I was thinking of different ways to help AIDS patients and I thought that a seven-day AM/PM pillbox would be useful. (I have used them, and I remember my grandmother used them for her multitude of pills.) Then at the beginning or end of every week, the patient or the caretaker can put all the pills (and half pills) in the proper compartments. Although this unfortunately excludes all the people who need to take syrups, I was talking with one of the patients who came in for a guide yesterday about adherence and she said it would help. She has a son in his teenage years and, while she gives him the pills most of the days, sometimes he goes out to play football (soccer) with his friends and forgets to take the pills with him. When I asked about how he would feel carrying around a pill box, and if he was comfortable taking it around his friends, she said that he didn’t mind and that some of his friends were on ARVs as well. I am getting a mixed view on exactly how prevalent or entrenched the stigma for AIDS really is. Driving down from Choebe this weekend, even in the smaller towns there were billboards about AIDS encouraging testing, a large ad for condoms, and a free condom box at the passport immigration. I also saw the first abstinence billboard, which seemed almost to contradict the “get tested” billboard on the opposite side of it.

that puts the effectiveness of the technology in serious doubt. Clearly, HDCT+BMT technology was not adequately assessed before entering widespread use, a failure that led to many women receiving a painful treatment that offered fewer benefits than initially believed.

Important vocabulary of technology assessment

or treatments as well as systems that aid in the delivery of healthcare. The technology can address any component along the healthcare continuum: prevention, screening, diagnosis, treatment, or rehabilitation. In this text you will see examples of technologies ranging from simple childhood vaccinations to prevent disease

In the rest of this chapter, we will outline in more detail the methods of technology assessment. They will guide our thinking as we examine new technologies through-

to complex total artificial hearts to treat end stage heart disease. While the complexity of the technology can differ dramatically, the process of healthcare technology assessment is the same. The ultimate goal of health technology assessment is

out the rest of the text. We begin with definitions of some important terms. Technology does not necessarily involve sophisticated or expensive James Bond like gadgetry or devices. Healthcare technology can be any intervention to promote health, including specific tests

to inform decision making, whether it is done from the perspective of an individual patient or from the larger perspective of society. The underlying questions which health technology assessment needs to address include the following [2].

Bioengineering and technology assessment

Prevention: health interventions designed to prevent a patient from developing disease. Screening: a test given to members of a defined population, not necessarily at risk for a disease, to identify those individuals who are most likely to be helped by further tests to diagnose the disease. Diagnosis: the identification of disease through signs, symptoms, imaging, bloodwork, cultures, cytologic sampling, or biopsy. Treatment: a health intervention to cure disease or to reduce symptoms of disease. Rehabilitation: the process of restoring skills lost to illness or injury.

r What is the clinical impact of the intervention? r What is the cost of the intervention? r What is the clinical impact of the intervention


Examples of benefits and costs of prostate cancer screening Monitoring the level of prostate specific antigen (PSA) is frequently used to screen older men to determine if they are likely to have prostate cancer. Invasive diagnostic testing is performed to determine whether men with elevated PSA levels have prostate cancer requiring therapy. Intended benefits: screening can identify men with prostate cancer at an early stage, when it is still curable. Intended costs: the cost of a PSA exam is less than $100. Unintended costs: because prostate cancer grows slowly, the new test can identify men with prostate cancer that will never cause any symptoms. These men undergo further invasive, painful testing and treatment which may be unnecessary [4].

weighed against its cost? In order to answer these questions, it is necessary to evaluate the safety, effectiveness, cost effectiveness, and the social, ethical, and legal impacts of a technology [3]. In this evaluation, we must consider both the direct and indirect consequences of using a health technology. Direct consequences of a health technology are the intended benefits and costs [3]. For example, if we develop a new, more accurate cancer screening test, the intended benefits include factors such as the accuracy of the test and the number of late stage cancers that could be prevented through early detection if the new test is widely adopted. The intended costs include the cost of the test, as well as the savings that would result from being able to treat patients for early stage cancer, which is less expensive than treatment for late stage cancer. In rare cases, the savings associated with a technology can actually be greater than the costs of using the technology. We will later see that this is the case with some childhood immunizations. The indirect consequences of a technology are the unintended economic, social, or other technology effects

[3]. Let’s imagine that our screening test is somewhat invasive and patients perceive it to be much more uncomfortable than the previous screening test. As a result, some patients who would have been screened with the old test now avoid cancer screening altogether, because of fear or embarrassment. In this case, the introduction of a new and more accurate test can actually decrease screening effectiveness because patient adherence to physician recommendations decreases. Health technology assessment must account for such unintended consequences. As such, health technology assessment is a bridge between the basic research and development of a technology and its real-life application. In the ideal world, technology assessment provides an opportunity to assess the technology’s effects before its widespread introduction, but in many cases it is used to analyze mature technologies already in routine clinical use to suggest strategies to use limited healthcare resources more effectively.


Biomedical Engineering for Global Health

When do we assess technology? A key question in health technology assessment regards the timing of when to perform a technology assessment. When health technology is assessed early, there are increased benefits, including potentially protecting public safety and identifying which populations should use the technology. However, assessment in the earlier stages also has risks. A health technology may not yet be perfected, and the populations for whom the technology should be used might not be appropriately identified. In addition, in an early assessment the data available about the performance of the technology are more likely to come from clinical trials rather than the settings where it will be routinely used. This can be a problem because clinical trials are often carried out by experts under well controlled conditions and may overestimate the performance of a technology in the community setting. We have a special vocabulary to describe this change in performance. Efficacy refers to the performance of a technology under ideal, controlled conditions. Efficacy is studied in homogeneous populations by using standardized procedures under ideal testing conditions by expert practitioners. Effectiveness is the measure of performance in a normal clinical setting. Effectiveness is studied in heterogeneous populations, and the technology is implemented by ordinary practitioners under conditions of routine clinical care [5]. Often the efficacy of a technology is much higher than its effectiveness, because the same experts that developed a test are using it under the best circumstances. If a health technology assessment is carried out using efficacy data, then the assessment lacks critical information about how the technology will truly perform in the real world. The true effectiveness, or impact, of the technology cannot be known until the technology has widespread use. The argument for later evaluation also has advantages and disadvantages. A later evaluation of technology will have more data, particularly regarding effectiveness, but the data may be biased. For example, if the technology has already been dispersed widely, then a randomized clinical trial may not be ethically feasible. If the initial effectiveness of the technology is favorable, researchers face an ethical dilemma by intentionally withholding the treatment to trial participants in the control group. Potential participants are also less likely

to accept randomization in such a trial if significant benefits have already been demonstrated. Additionally, by the time a health technology assessment is completed the technology may already be outdated, either in how it is specifically used, or because superior alternatives for the given clinical problem have been identified.

Technology assessment in developing countries Technology assessment is particularly important in settings where healthcare resources are extremely limited, such as developing countries. Unfortunately, few developing countries have health technology assessment programs. The results of technology assessment applied in different settings cannot simply be used in developing countries – many factors, such as whether a disease is common or rare, the social acceptability of a technology, the efficacy of a technology and the cost – vary dramatically throughout the world [6].

Thus, the timing of healthcare technology assessment is a “moving target” problem with no obvious answer. It is for this reason that health technology assessment should be an iterative process, done on an ongoing basis to achieve evaluations throughout the life cycle of the technology development process.

Metrics of health technology assessment Many times, a ratio of benefit to cost will be calculated as a quantitative metric of technology assessment. It is important to point out that the decision regarding the usefulness of a technology cannot be made in a vacuum. In other words, it must be recognized that, when health technology is assessed, one must think about the relative benefits and costs involved in the clinical situation. There is always an alternative to a new health technology being assessed. Alternatives include currently used treatments or technologies, termed the standard of care. Both the standard of care and the new technology have economic costs and clinical benefits that must be taken into account. Ranking strategies according to their benefit to cost ratio can often be a helpful way

Bioengineering and technology assessment


to compare the effectiveness of different approaches, including new technologies, the standard of care and “do nothing” strategies [7].

tal group receiving the new technology and a control group receiving existing technology, should be used to compare the performance of the new technology to the existing technology. In the hierarchy of information,

Collecting data for health technology assessment

randomized clinical trials are considered the strongest study design [8]. Randomized clinical trials may not be necessary if the patient benefit from the test is so dramatic as to leave little room for doubt that the new technology is as good or better and less expensive than

Where do we obtain data about the costs and benefits of healthcare technologies? Generally data come from clinical trials – well controlled experiments designed to compare the performance of two technologies. In carrying out health technology assessment, we can obtain secondary data about a technology from published literature describing clinical trials, or we can collect primary data by carrying out our own clinical trial. Primary data can be analyzed for efficacy, effectiveness, safety, reproducibility, patient satisfaction, and cost effectiveness. Secondary data can also be used to assess the same outcomes – either by using data from one published study or by using meta-analysis, a statistical method that combines data from different studies to estimate the overall effect of an intervention on a specific outcome [5]. As we will see throughout this text, carrying out a clinical trial can be an expensive task; if we want to examine long term outcomes, we may have to wait decades for results. Many times we want to carry out a health technology assessment without waiting this long. As an alternative, we can write a computer program to simulate the clinical trial. This process is called decision analysis. In decision analysis, we simulate the likely outcomes from a group of hypothetical patients, using probabilistic methods [5]. Data from primary and secondary sources are used to estimate the efficiency of a technology as well as likely patient outcomes. The computer program follows each group of patients over time, using these data to “roll the dice” for each patient at important time points. Decision analysis can provide a very quick and inexpensive way to estimate what might happen in a clinical trial without having to spend millions of dollars and wait decades. However, it is not a substitute for actually carrying out a clinical trial. There are many types of clinical trials that can provide useful data for health technology assessment. Nonrandomized clinical series often provide data on the efficacy of a technology. Randomized clinical trials, where participants are randomly divided into an experimen-

existing options. However, few technologies meet those criteria.

Policy decisions and HTA How is a health technology identified as a candidate for assessment? How do the results of health technology assessment affect the use of a technology? Who monitors health technology assessments? These are all crucial policy questions that are just as important as the scientific methodology of health technology assessment. As mentioned earlier, the purpose of health technology assessment is to aid and inform decision making regarding the use of a technology. It is important then to understand not only how a technology is assessed, but also what is then done with that assessment.

Consensus conference clinical guidelines The National Institutes of Health (NIH) is the medical research agency of the United States. One function of the NIH is to organize conferences to bring together expert scientists and physicians to produce consensus statements on important and controversial topics in medicine. Consensus recommendations are based on publicly available scientific data. These consensus guidelines influence the practice of many physicians throughout the world. For example, the 2000 NIH Consensus Statement on Adjuvant Therapy for Breast Cancer recommends that the majority of women who have localized breast cancer be treated both with surgery and also with chemotherapy because of the small but statistically significant improvement in survival [9].


Biomedical Engineering for Global Health

Often emerging technologies surface through reports in the literature of a series of cases. Once an existing technology has been accepted as standard clini-

well designed, well conducted studies in representative

cal practice, its use can be tracked with data obtained from health registers and institutional and organizational databases, and by use of national administrative and financial data and post-marketing surveillance data. Often policy decisions about the use of existing technologies are then made by group judgment methods (e.g. consensus conference) [9]. Social and ethical issues should be considered throughout the develop-

effects on health outcomes, the strength of the evidence may be limited by number, quality, or consistency of the studies, Poor evidence is considered insufficient to

ment of a technology.

Clinical preventive services guidelines Recommendation on screening for HIV infection Clinicians should assess risk factors for HIV infection by obtaining a careful sexual history and inquiring about injection drug use in all patients. Periodic screening for infection with HIV is recommended for all adolescents and adults at increased risk of infection. Early therapeutic intervention reduces the risk of clinical progression and mortality. Screening is recommended for all pregnant women. Treatment can significantly reduce rates of mother-to-child transmission. The US Preventive Services Task Force makes no

populations that directly assess effects on health outcomes. While fair evidence is sufficient to determine

assess the effects on health outcomes due to a limited number of studies or flaws in the study design. The strength of recommendation is divided into five categories: there is good, fair, or insufficient evidence to support the recommendation that the test not be used in periodic health examinations, and there is fair or good evidence to support the recommendation that the test be used in periodic health examinations. This hierarchy has been applied by many technology assessors. Not all procedures apply to the periodic health examination, but the concept of using evidence to justify the strength of a recommendation is logical.

Learn more about the Institute of Medicine Health Care Quality Initiative The report Crossing the Quality Chasm: A New Health System For The 21st Century was issued by the IOM in 2001. This report calls for comprehensive reform of the healthcare system to ensure that all patients receive quality, evidence

recommendation for or against routinely screening non-pregnant adolescents and adults who are not at increased risk for HIV infection. The benefits of screening those without risk factors are too small relative to potential harms to justify a recommendation [4].

based care [12]. Committee on Quality of Healthcare in America. (2001). Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century,

The US and Canadian Preventive Services Task Forces, which make recommendations about clinical preventive services, have devised a hierarchy of evidence upon which to base recommendations. The hier-

The most recent Clinical Preventive Services Guidelines include recommendations for several tests we will examine later in this book [4]. For example, the current recommendation regarding breast cancer screening sug-

archy has two determinations: quality of evidence and strength of recommendation [4]. The Task Forces categorize the overall quality on a three point scale divided into good, fair and poor. Good evidence is derived from

gests that all women aged 40 and over undergo a mammogram every one–two years, with or without clinical breast examination. In contrast, screening for HIV is not universally recommended. According to the guide, only

Washington, D.C.: National Academy Press.

Bioengineering and technology assessment pregnant women and non-pregnant adults and adolescents who are at risk for HIV need to be regularly tested. All recommendations are made by reviewing the available evidence; a test, service, or immunization is recommended only when the data suggests that it will be effective. Given that health technology assessment has an important effect on policy, it is important to track its use and application, and to ensure that health services delivered to patients are consistent with current professional knowledge. The Institute of Medicine (IOM) has undertaken a comprehensive effort to assess and improve the quality of healthcare throughout the United States. The first phase of this initiative began in 1996, and documented serious problems of the quality of healthcare delivered in the United States, concluding that the burden of harm conveyed by healthcare quality problems is


tal, at the level of healthcare organizations, and at the interface between clinicians and patients. As we implement reforms to improve the quality of health systems, it is important to measure whether these reforms actually improve the health of our population. In the next chapter, we will look at various types of health data which are used to assess health status. We will use these measures to compare health status of populations throughout the world.

Homework 1. Two scientists want to know if a certain drug is effective against high blood pressure. The first scientist wants to give the drug to 1000 people with high blood pressure and see how many experience lower blood pressure levels. The second wants to

r Only 55% of patients in the USA receive care consis-

give the drug to 500 people with high blood pressure and not give the drug to another 500 people with high blood pressure and see how many people in both groups experience lower blood

tent with consensus guidelines. r The delay between the discovery of more effective

pressure. Source: [13]. a. What is the better way to test this drug?

staggering [10]. The following are some examples [11].

forms of treatment and the incorporation of these treatments into routine patient care averages 17 years. r More than 18,000 Americans die every year as a result of heart attacks because they did not receive preventive medications, even though they were eligible to receive them. r More Americans are killed every year as a result of medical errors than by breast cancer, AIDS or motor vehicle accidents. In the second phase of the review, the IOM established a vision to transform the US healthcare system in order to close the gap between quality care and what exists today in practice. Recommendations include establishing healthcare systems where decision making is evidence based rather than based on a physician’s training and experience, and shifting the view that patient safety is ensured by an individual’s responsibility to “do no harm,” to one where safety is an inherent property of the healthcare system as a whole [12]. The third phase of the review, currently ongoing, is focused at implementing these reforms on three levels, the environmen-

b. Why is it better to test the drug this way? 2. Find a news report describing a new health technology published in the past year. a. Based on this article, summarize which steps in the technology assessment process have been carried out for this technology. b. Given this, do you believe that the news report provides a balanced discussion of the potential promise and the potential limitations of this technology? 3. Dr. Maurice Hilleman died recently. A quote from his obituary stated, “I think it can be said without hyperbole that he was a scientist who saved more lives than any other modern scientist.” a. What was Dr. Hilleman’s contribution to medical science? b. How many lives per year are saved as a result of his work? c. Discuss Dr. Hilleman’s work as an example of translational research. n29/obit.html


Biomedical Engineering for Global Health

References [1] Littenberg B. Technology assessment in medicine. Academic Medicine: Journal of The Association of American Medical Colleges. 1992 Jul; 67(7): 424–8. [2] Deber RB. Translating technology assessment into policy. Conceptual issues and tough choices. International Journal of Technology Assessment in Health Care. 1992 Winter; 8(1): 131–7. [3] Szczepura A, Kankaanpaa J. An Introduction to Health Technology Assessment. In: Szczepura A, Kankaanpaa J, eds. Assessment of Health Care Technologies. New York: John Wiley & Sons; 1996. [4] Guide to Clinical Preventive Services, 2006. June 2006 [cited AHRQ Publication No. 06–0588]; Available from: [5] Goodman C. A basic methodology toolkit. In: Szczepura A, Kankaanpaa J, eds. Assessment of Health Care Technologies. New York: John Wiley & Sons; 1996. [6] Tan-Torres T. Technology assessment in developing countries. World Health Forum. 1995; 16(1): 74–6. [7] Cantor SB, Ganiats TG. Incremental cost-effectiveness analysis: the optimal strategy depends on the strategy


[9] [10]




set. Journal of Clinical Epidemiology. 1999 Jun; 52(6): 517–22. U.S. Preventive Services Task Force. Guide to Clinical Preventive Services. 2nd edn. Alexandria, VA: International Medical Publishing; 1996. Adjuvant Therapy for Breast Cancer. NIH Consensus Statement. 2000 November 1–3; 17(4): 1–35. Chassin MR, Galvin RW. The urgent need to improve health care quality. Institute of Medicine National Roundtable on Health Care Quality. JAMA: The Journal of The American Medical Association. 1998 Sep 16; 280(11): 1000–5. Institute of Medicine. The Chasm in Quality: Select Indicators from Recent Reports. 2006 May 30 [cited 2007 May 28]; Available from: Institute of Medicine (U.S.). Committee on Quality of Health Care in America. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, D.C.: National Academy Press; 2001. 337 pages. National Science Board. 2008. Science and Engineering Indicators 2008. Two volumes. Arlington, VA: National Science Foundation (volume 1, NSB 08-01; volume 2, NSB 08-01A).

3 Health and economic data: a global comparison

As we consider the development of new technologies to improve health, it is important to step back and consider how we define health and how we assess the health of a population. We all have our own perceptions of disease; things like pain, fever, and symptoms of illness which can interfere with our normal activities, or reduce our ability to respond to stress and physical injury (Figure 3.1). The World Health Organization defines health to be “a state of complete physical, mental and social well being and not merely the absence of disease or infirmity [1].” In Chapters 3 and 4, we will see that the health status of a population is frequently correlated with economic measures such as income and health expenditures. The focus of this book is the engineering of new health technologies to meet world health needs. The goal of bioengineering design is to apply advances in science to solve health problems in a way that meets resource constraints. Thus, to engineer new technologies we must understand both global health needs and global resource constraints. Both health needs and resource constraints vary dramatically throughout the world – a technology that solves a health problem in one part of the world, may not be a solution in a different part of the world. In this chapter, we will develop metrics to assess the health and economic status of populations. We will use

Figure 3.1. A mother and her son smile after doctors at Chicuque Hospital in Mozambique saved his life. Absence of disease is only one component of overall health. Shannon Trilli, 2003. Used by permission of UMCOR (United Methodist Committee on Relief).


Biomedical Engineering for Global Health

these metrics to compare the health of different regions, as well as the ability of new technologies to improve health throughout the world. This will help give us a clear picture of where technologies have been effective, and where new technologies are needed.

Health data How does the health of a population differ from an individual’s health? In characterizing the health of a population, we must somehow pool together data about the health of the individuals that comprise that population. Scientists called epidemiologists specialize in the study of the health of populations. They calculate pooled figures such as infant mortality rates, numbers of deaths and causes, and immunization rates to develop a picture of the health of a population. In this chapter, we will examine why we need health data, what data we need, where we obtain these data, and how we use them to improve health. The importance of health data became clear at the beginning of the twentieth century. From 1870 to 1900, biomedical science advanced more than it had in the previous three millennia. In this period, Darwin’s concept of evolution was established; chemistry and microscopy were used to carry out field based research around the world. During this time, the means, transmission route, and causative agent of almost every important infectious disease were established. With this

Portrait of epidemiologist John Snow In the 1840s, cholera outbreaks in London claimed many lives. At the time, most scientists believed that cholera was transmitted by breathing contaminated vapors. John Snow believed that cholera was spread through contaminated food and water but was unable to prove this theory. In 1854, a cholera outbreak struck London. Snow began plotting the number of deaths by location throughout the city. At this time, London received its water supply from two companies, one which drew water from the Thames upstream of the city and the other which drew water downstream of the city. Snow noted that the concentration of cholera victims was higher in areas of the city supplied by water drawn downstream of the city, a location that was more likely to be contaminated by city sewage. Snow noticed a particularly high number of cases at the intersection of Cambridge and Broad streets – more than 500 people died of cholera over a ten day period. Snow convinced city officials to remove the pump handle that supplied water to this neighborhood and the epidemic was contained. Snow’s study was one of the first epidemiologic analyses, and it firmly established the value of health data in tracking and eliminating the spread of disease [2].

improved understanding of disease and how to control it, governmental health agencies were first established. The World Health Organization (WHO) was established by charter of the United Nations after World War II. The WHO is headquartered in Geneva, and its mission is the “attainment by all peoples of the highest possible level of health [1].” The WHO serves a critical role, providing important health information to governments, including epidemiologic intelligence and data on world health problems, international standardization of vaccines, and reports of expert committees on health problems (Figure 3.2). Countries that are members of the WHO must provide certain information in regular reports, including information about disease outbreaks, the health of their population, and steps the country is taking to improve health. The WHO’s website

Source: Snow, J. On the Mode of Communication of Cholera. 1855.

Health and economic data: a global comparison


Figure 3.2. This map displays locations of documented malaria outbreaks. The WHO compiles data on a range of health related issues, such as malaria, ensuring that crucial health concerns are addressed and resources allocated effectively [6].

( includes much useful health data that we will refer to throughout the course. How do societies use this type of health data? Health data can be used to provide an early warning system to identify emerging health problems. For instance, epidemiological data helped established the connection between rubella early in pregnancy and severe birth defects, even fetal death. As a result, rubella vaccines are now routinely administered during childhood, and cases of rubella associated birth defects have decreased. Another example involves thalidomide, a drug once used to treat morning sickness. Beginning in late 1959, cases of rare, severe birth defects involving the limbs and digits started to accumulate in high numbers (Figure 3.3). Given the sudden spike in the number of cases, investigators suspected a new drug might be the culprit; soon after, the defects were traced to thalido-

Figure 3.3. Analysis of health statistics led to the removal of thalidomide from the market after it proved to be very harmful for pregnant women, as shown in this photo of a baby born with an extra appendage on the foot. Source: NCI/G. Terry Sharrer, Ph.D. National Museum of American History.

mide and in 1961 the drug was taken off the market [7]. In 1981, two diseases almost exclusively seen in older, immunocompromised people, Kaposi’s sarcoma and Pneumocystis carinii pneumonia (PCP), began to appear in young, previously healthy, homosexual men [8]. As a result of these unusual appearances, the CDC launched an investigation into what appeared to be a new disease, later identified as AIDS. Health data can be used to help estimate the impact of health problems, such as the number of people affected by a disease and their ages and locations, to determine public policy to respond to a disease, to educate legislators and set


Biomedical Engineering for Global Health

societal priorities for healthcare funding, and to monitor progress toward goals so that interventions can be assessed objectively. Given these uses, what types of health data should be collected? Health data can include data on the population of a country, such as the number of people, their age, sex, ethnic origin, and urbanization. They include vital statistics, such as the number of live births, and the number of deaths (including infant deaths) by sex, age, and cause. Such data are essential in order to accurately assess the health of populations and make decisions regarding health resources; unfortunately these data are lacking from many areas. Vital statistic data from many countries are not complete. In some countries deaths are recorded only in certain areas, while in others all areas are covered but not all deaths are recorded. Average rates of coverage vary widely, from only 10% in Africa, to over 90% in Europe [9].

reported to health agencies. For example, the WHO maintains a list of reportable diseases. This allows for monitoring of potential outbreaks and prevention of epidemic spread. In the United States the Centers for Disease Control and Prevention (CDC) in Atlanta manages mandatory reporting to the WHO. The CDC also maintains a list of nationally notifiable diseases [10]. State health departments collaborate with the CDC to determine which diseases should be on this list, and the list varies slightly from state to state. State reporting is voluntary, although all states report diseases in compliance with WHO regulations. In addition to tracking reportable diseases, the WHO also maintains tumor registries, which compile epidemiological data regarding cancer cases. A final category of health data tracks health services, including the number and type of healthcare facilities, the number and qualifications of health personnel, health services and utilization rates, and costs and payment mechanisms.

World Health Report Each year, the WHO publishes the World Health Report. This report provides data quantifying health throughout the world, tracks progress in improving world health and gives an overview of a specific health problem. The topics of past World Health Reports include the following. 2006: Working Together for Health calls for improvements in the health workforce throughout the world [3]. 2005: Make Every Mother and Child Count calls for ways to improve maternal and child health throughout the world [4]. 2004: Changing History calls for a comprehensive HIV/AIDS strategy [5].

Health data also include health statistics, such as the frequency of disease by type, severity and outcome. Certain epidemic-prone diseases are considered to be “reportable” diseases, meaning any incidence must be

Annual incidence rate =

Nationally notifiable diseases The 2007 list of US nationally notifiable diseases includes 63 infectious diseases, including AIDS, anthrax, mumps, pertussis, plague, rubella, smallpox, tuberculosis, and typhoid fever [10]. The complete list can be found at:

Quantitative health measures If we are to develop an accurate picture of health throughout the world, it is important to measure such health data in a quantitative manner. A number of health statistics have been found to be useful in assessing the health of a population. Here we examine some of the most important. Incidence refers to the number of new cases of a disease in a population over a period of time. We can calculate the annual incidence rate of a disease as follows [7].

Number of new cases of a defined condition in a defined population in one year Number in that population at mid-year of that same year


Health and economic data: a global comparison


In contrast, the prevalence of a disease indicates the number of existing cases of the disease in a given population at a specific time. The prevalence of a disease is

and prevalence of disease in populations? These data allow one to estimate the magnitude of health problems and to detect epidemics; one example is the “reportable

calculated as follows [7].

disease” law discussed earlier. In 1951 the WHO

Point Prevalence =

Number of cases of a defined condition in a defined population at a point in time Number in that population at same point in time

While the definitions of incidence and prevalence appear similar, they are actually quite different and it is important to appreciate the distinction. When would incidence and prevalence of a disease differ? Consider a disease with a relatively short duration such as the flu – the annual incidence rate is much higher than the point prevalence, because while many people contract the flu each year, at any given time throughout the year they are not all sick. In contrast, for a disease with a relatively long duration, such as HIV/AIDS, the point prevalence can be much higher than the annual incidence rate. Why is it important to examine incidence


adopted the International Health Regulations (IHR), global legislation that requires all countries to notify the WHO of incidences of specific “quarantinable diseases.” The first new case must be reported within twenty-four hours, all subsequent cases and deaths must be reported as well. This mandatory notification is currently required for cases of cholera, plague, and yellow fever [11]. However, owing to the renewed spread of old diseases and the rise of new ones, the WHO is currently working to broaden the powers of the IHR to cover any “public health emergency of international concern [12].”

Global outbreak alert and action In 2000, the WHO formed a new infrastructure to organize world response to outbreaks of infectious disease. The Global Outbreak Alert and Response Network (GOARN) unites 130 existing agencies and networks throughout the world to work together to systematically detect and verify disease outbreaks, to provide real time alerts and to respond rapidly to contain outbreaks [13]. The effectiveness of GOARN relies on international cooperation. More than 800 people died in the SARS outbreak in 2002–3. World response to the outbreak was delayed because Chinese officials delayed full disclosure of initial cases of SARS. Courtesy of WHO/Global Outbreak Alert and Response network


Biomedical Engineering for Global Health

Once a potential outbreak is identified, preventative measures can be put in place to avert an epidemic. These include sending supplies and expert personnel

publishes the Weekly Epidemiological Record (WER); an essential tool for providing health personnel with information pertaining to outbreaks of communicable

to the disease site, as well as offering technical advice and sometimes initiating an epidemiological investigation. To facilitate information dissemination the WHO

diseases [14]. Likewise the CDC publishes a morbidity and mortality weekly report (MMWR), available at

Health and economic data: a global comparison The mortality rate in a population quantifies how many people have died. Many types of mortality rates are monitored; however, we will examine two in this text. The first is the crude death rate, often referred to as the mortality rate. The mortality rate is defined as follows.

Mortality Rate =


in organizing these data to create an overall picture of world health, identifying which diseases are most prevalent, where outbreaks are occurring, and how specific health concerns can be addressed. First, health data can suggest which diseases should be cause for great concern. For example, leading causes of mortality

Number of deaths in a defined population in a year Number in that population at mid-year of the same year

Infant mortality refers to the number of deaths of persons under one year of age and is defined as follows.

Infant Mortality Rate =


throughout the world help prioritize which diseases constitute the greatest threat to world health, allowing

Number of deaths under 1 yr of age in a defined population in a year Number of live births in that population in same year

The infant mortality rate is often used as an indicator of how well a country’s health system functions. Finally, health data can be used to estimate not only the occurrence of disease and death, but also the burden of disease. Morbidity refers to the degree or severity of a disease. We will use one measure of the burden of disease which combines the effects of both morbidity and mortality, the disability adjusted life year (DALY). DALYs measure the years of disability free life lost when a person contracts a disease. A DALY combines several elements, including the levels of mortality by age, the levels of morbidity by age and the value of a year of life at specific ages. For example, it is assumed that losing a year of life at age 20 is more serious than losing a year of life at age 90. While we will not learn to calculate DALYs (a complex calculation), we will use the DALY to compare the impact that different diseases have on different populations. Table 3.1 provides the DALYs lost per person associated with some common diseases and conditions in two regions of the world. In each case, you can think of a DALY as the average number of years of disability free life that an individual who contracts that disease or condition would lose.

Organizing and interpreting health data The WHO collects vast quantities of data yearly regarding a multitude of health statistics. The challenge lies


resources to be targeted effectively. HIV/AIDS, heart disease, and cancer are all leading causes of death (Table 3.2), and correspondingly many resources are devoted to improving their prevention, diagnosis, and treatment. [16] Health data are also commonly used to learn more about the global impact of a particular disease. In 2004, the WHO devoted its entire annual report to the HIV/AIDS pandemic [5]. Data showed that although AIDS affects people throughout the world it does not do so equally; roughly 2/3 of those affected live in Africa (Figure 3.4). In order to effectively address HIV/AIDS, the WHO has called for an acceleration of

Table 3.1. Average DALYs lost per person in North America and Africa for some common diseases and conditions [15]. Disease/Condition

DALYs Lost /Person North America





Car accidents



Self-inflicted injuries










Lower respiratory infections HIV


Biomedical Engineering for Global Health

Table 3.2. Leading causes of death by age [16]. Leading causes of mortality among adults worldwide 2002 Age: 15–59 Rank

Age: 60+


Deaths (000)





Ischemic heart disease





Deaths (000)


Ischemic heart disease




Cerebrovascular disease




Chronic obstructive pulmonary disease


Road traffic injuries



Lower respiratory infections



Cerebrovascular disease



Trachea, bronchus, lung cancers



Self-inflicted injuries



Diabetes mellitus






Hypertensive heart disease



Cirrhosis of the liver



Stomach cancer



Lower respiratory infections





Chronic obstructive pulmonary disease



Colon and rectum cancers




40 35 30

25 Number of people 20 living with HIV 15


Oceania Middle East & North Africa Eastern Europe & Central Asia Latin America and Caribbean North America and W & C Europe Asia Sub-Saharan Africa

10 5 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Year Figure 3.4. Estimated number of adults and children living with HIV, by region, 1990–2007. Reproduced by kind permission of UNAIDS.

both prevention and treatment programs in those areas hit hardest by the disease. Health data can identify how to address health problems. Vital statistics can be used to track birth rates and identify areas that need improved prenatal care or other measures to prevent infant mortality. For example, Table 3.3 lists the leading causes of infant mor-

tality in the developing world. Three of seven are diseases that could be prevented with a simple vaccination: measles, pertussis (whooping cough), and tetanus. These data suggest that increased use of childhood vaccinations could have dramatic effects on infant mortality, and, motivated by this need, researchers are currently working to develop new technologies

Health and economic data: a global comparison Table 3.3. Leading causes of infant mortality in the developing world [16]. Causes

Numbers (000)

Lower respiratory infections


Diarrhoeal diseases













to make it easier to transport and administer these vaccines. Finally, health data can help identify the factors that contribute to morbidity and mortality. Epidemiologic studies help identify those factors which are associated with disease. Association is simply the statistical dependence between two or more events (e.g. the incidence of cancer rises with increasing age); association does not imply that one event causes the other [7]. However, understanding how certain factors are associated with health status can help us generate hypotheses about cause and effect, and can help us identify populations at most risk. A metric frequently used to characterize the strength of an association is relative risk. The relative risk is simply the risk of disease in a population that has been exposed to a certain factor divided by the risk

New Adherence Project . . . : July 12, 2007 Tessa Swaziland Now that the new volunteers have taken over most of the World Food Programme (WFP) duties, I’ve been free to work on the new adherence project with Tina. Some background info about outreach . . . Baylor sends doctors to Good Shepherd roll-out clinics. These are clinics that are located in rural villages and treat mild ailments. But once a month, Good Shepherd (the hospital in the Lobamba region of Swaziland) sends nurses to distribute to the HIV patients there. The Baylor doctors accompany them to see the patients, assess their health, and assess their drug regimen. Good Shepherd uses a certain form to record the adherence of each patient. It basically has a place for counting the pills and then a place for a general assessment, but the adherence percentage is never calculated. Also, no recommendations concerning the adherence level are made. Tina gave me a new form to edit and test out in the field. This form makes it easy to calculate the adherence percent and make an appropriate recommendation (ie, continue ART, see an adherence counselor, refer to doctor, put in “high-risk adherence failure” group, etc.). Before I went to the outreach clinics, I edited the sheet to incorporate the system they use now. I figured the fewer changes they had to make, the more likely they would be to change. The first day we tried it out was Tuesday. Tina explained to the Good Shepherd adherence counselors what I was doing. One of them was happy to help me out, but the other one was initially very resistant. So I spent the first day filling the sheets out myself alongside them. This gave me a chance to get to know the new counselor better. At the end of the day, they looked at the sheets with me and gave me input about what they thought worked well and what needed to be changed. Today, only one of the counselor was there. She was much more receptive and friendly, and she actually had to use the forms for the last ten or so patients since she ran out of her own forms. That gave her a chance to actually see what she thought about the new adherence sheet. Once she got the percentage calculations down, she seemed very pleased and asked for extra copies to take back with her to the hospital.


Biomedical Engineering for Global Health

Table 3.4. Relative risk of death associated with exposure to several factors [17–20]. Groups Compared

Relative Risk of Death

Death from Lung Cancer: People who smoke versus People who do not smoke

RR = 15

Death from SIDS: Infants of mothers who smoke versus Infants of mothers who do not smoke

RR = 5

Death from Diarrheal Disease: Infants 99% [14]. Flow cytometry is an important tool for monitoring patients with HIV. HIV infects and kills certain types of white blood cells called CD4 lymphocytes. The number of CD4 lymphocytes (CD4 count) is critical to determine the clinical stage of HIV infection, to evaluate whether treatment is working and to determine when medications need to be changed. The CD4 count is measured in a flow cytometer. White blood cells are stained with a fluorescent antibody targeted against the CD4 surface marker in order to quantify the number of CD4 cells present in clinical samples. The cost of a flow


Biomedical Engineering for Global Health



(b) (b)

Figure 7.17. (a) Schematic of how a flow cytometer works. Reprinted with permission of John Wiley & Sons, Inc. Flow Cytometry: c 1992 Wiley-Liss, Inc. First Principles; Alice L. Givan; Copyright  (b) A flow cytometry system. Courtesy of the Center for Flow Cytometry, Czech Republic.

cytometer ranges from about $30,000 to $150,000; they are not available in many low-resource settings because of their high cost. As a result, many patients with HIV or AIDS do not currently have access to this important test [15]. Advances in microelectronics technology provide an exciting opportunity to reduce the cost of high throughput biosensors. Using microfabrication techniques, a number of lab-on-a-chip systems have been developed to carry out chemical analyses with pocket sized equipment. McDevitt and colleagues at the University of Texas have developed a microchip to rapidly quantify

Figure 7.18. (a) Whole blood sample processed through flow cell; (b) image of fluorescently stained CD4 lymphocytes. Source: Rodriguez et al. Plos Med. 2005. Z (7): el82.

CD4 lymphocytes at substantially reduced cost compared to flow cytometry [15]. Whole blood from the patient is introduced into a small flow cell; white blood cells are captured on a membrane which excludes red blood cells. A fluorescent antibody is used to stain CD4 lymphocytes, which are imaged using a color camera. Image processing algorithms are used to automatically identify and count the number of CD4 lymphocytes present. Such sensors may have great applicability to improve healthcare in the developing world. In low resource settings, provision of laboratory services is frequently difficult because there may be limited access to running water or electricity and ambient temperature and

The evolution of technology humidity can fluctuate widely. In addition, consumable reagents required for diagnostic tests may frequently be unavailable. Even when laboratory facilities are available, there is often a lack of trained laboratory personnel in many developing countries [16]. Thus, there is an important need for simple, low cost techniques to perform diagnostic tests such as blood chemistries, immunoassays, and flow cytometry in low resource settings. Disposable immunoassay tests, which use inexpensive components, can be mass produced, and are relatively affordable, have been successfully used in many developing countries. A disposable immunoassay test consists of a nitrocellulose membrane strip containing all of the dried reagents necessary to test for the presence of an antigen in a small amount of liquid sam-

(a) Absorbent pad Positive test result

is present, then no color change occurs at the test line. As the solution passes the control line, some labeled

Negative test result Plastic-backed nitrocellulose membrane

(b) Lysing agent. Labled Ab.

Test band (bound Ab).

Control band (bound Ab).

Bound Antibody Nitrocellulose strip. Free labled Ab.


Buffer/ flushing agent.

ies which will bind the antibody present on the surface of the spheres in the sample pad (Figure 7.19) [16]. In order to perform the test, the liquid sample is

present, some of the complexes of antigen and labeled antibodies will bind to the immobilized antibody at the test line; essentially, the antigen acts as a “sandwich” to link the immobilized test antibody and the labeled antibody. The aggregation of spheres results in a color change at the test line location. If no antigen

Test line

Sample pad

test contains three different regions: (1) the sample pad contains tiny spheres made of gold or colored latex which are coated with antibodies that bind to the target antigen to be detected, (2) the test line contains

applied to the sample pad at the end of the test strip. This solubilizes the colored spheres coated with antibodies stored on the test strip. If the antigen is present in the sample, it binds to the solubilized antibody present on the surface of these colored spheres. The solution moves through the membrane by capillary action. As the solution passes the test line, a color change will occur only if the target antigen is present. If antigen is

Control line

Colloidal goldantigen conjugate

ple solution (usually urine, blood or saliva). The strip

a line of physically immobilized antibodies which will bind to the antigen to be tested, and (3) the control line contains a line of physically immobilized antibod-


Parasitized Blood Parasite antigen (Ag.) captured by labled Ab.

Blood and labled Ab flushed along strip.

(d) Captured Ag-labled Ab complex

Labled Ab-Ag complex captured by bound Ab of test band.

Captured labled Ab.

Labled Ab captured by bound Ab of control band

Figure 7.19. (a) A disposable immunnoassay test, or strip test. Reprinted by permission from Macmillan Publishers Ltd: Nature 442: c 2006. (b)–(d) Mechanism behind malaria rapid 412–18,  diagnostic test. Courtesy of WHO,


Biomedical Engineering for Global Health

MCH Lab: June 26, 2007 Kim Malawi The district hospital lab was an interesting experience. I got several good ideas for possible design projects, and I learned a lot about the limitations that they’re working under here. Tests that we take for granted in the States aren’t possible here on a regular basis, and some are not possible at all. The blood chemistry analyzer is out of order (and has been for months) so all of those kinds of tests are impossible. No cell counts. No enzyme levels. No basic diagnostic tests like those! The automatic hemoglobin reader was also out of order. They were doing hemoglobin measurements by taking a hematocrit (they fill a capillary tube with blood, spin it in this special centrifuge, then use a device with an arm that you point at the division of plasma and blood and it gives you the packed cell percent. They then divide that number by three to get the Hb.) They were using a glucometer (like the over-the-counter ones in the States for diabetics) to do blood and CSF glucose levels. For now, this is working. But when they run out of the proprietary test strips, they’ll be out of luck on glucose tests. They don’t have a histology department anymore because the pathologist left. They do the “heat until the fluid begins to vaporize” step of the TB stain procedure by lighting a piece of cotton wool on fire and holding it with tongs. So not safe. (I watched the guy nearly light his sleeve.) There is one person who spends his entire day reading malaria blood smears. (Most are negative, as it turns out.) I’m interested now in seeing a rural health center (if I can) when we go out to Rhumpi or Chitipa, because those are where the bulk of health care in Malawi actually happens and Ellie says they are woefully underfunded and understaffed. She says it will be even less capable of conducting basic tests than the MCH lab. What a thought.

antibodies directly bind to the immobilized capture antibodies; the aggregation of spheres results in a color

In some cases it is desirable to quantify the amount of antigen present (e.g. malaria) and in these cases a

change. A color change at the control line indicates that reagent has passed this line and confirms that the antibodies present on the test strip are functional. Results are usually available in 5–15 minutes, and the test is read by looking for a change in color at the test line; quality control is ensured by looking for a change in

test which provides a quantitative or semi-quantitative result is necessary. To address this need, many researchers are developing point-of-care diagnostic systems that use a disposable card in conjunction with a low cost reader apparatus [16]. The sample is applied to the disposable card, which contains any necessary con-

color at the control line. Test sensitivity can be very high. For example, rapid diagnostic tests for the hepatitis B surface antigen can measure as little as 1.0 ng of antigen per ml of blood [17]. Typically, test strips are

sumable reagents and calibration supplies, and retains waste materials; a quantitative result is obtained by inserting the card into the reader. Figure 7.20 shows an example of a prototype device to

stable for months when properly protected from moisture and excessive heat. Such tests are routinely used to screen individuals for HIV infection in many developing countries. These disposable tests provide a yes/no answer and the test can be performed accurately by personnel with minimal training.

measure small molecule analytes and drug metabolites in saliva developed at the University of Washington. Filtered saliva is placed at the sample port; the sample is transported through a series of microfluidic channels where it is separated, labeled with antibody, and transported to assay channels, which are then interpreted

The evolution of technology


ucts. Complete analysis takes less than 30 minutes and the cost is between $1 and $5 per disposable test [16]. The field of bioinstrumentation offers many opportunities to develop new medical sensors and implants, and devices to improve laboratory diagnostics in a wide variety of settings. Usually, new biomedical devices and sensors are developed by teams of bioengineers, electrical engineers, neuroscientists, materials scientists and chemists, working together with clinicians or specialists in laboratory medicine who understand the design requirements and clinical needs of their field. Such multi-disciplinary collaboration can rapidly lead to new prototype devices for clinical testing.

Biomechanics The cells and tissues in our bodies are continuously exposed to a wide variety of mechanical forces. When we walk, muscles exert tensile forces which are transmitted through tendons to act on bones and move joints. Walking and running generate substantial compressive loads on cartilage and bones. As our heart pumps blood, changes in blood pressure generate hoop stress which causes blood vessels to dilate cyclically. The frictional forces associated with blood flowing past the vessel wall Figure 7.20. Lab-on-a-chip technology using microfluidics to rapidly detect various small molecules and metabolites. Reprinted by permission from Macmillan Publishers Ltd: Nature, 442: 412–18,  c 2006.

in the reader [16]. A similar approach has been used to develop a point-of-care diagnostic tool to identify pathogens responsible for enteric infections, which kill more than three million children each year, mostly in the developing world [16]. Tools used in developed countries (stool culture, enzyme immunoassay, and PCR), are not available in most laboratories in developing countries. As an alternative, a disposable card has been developed to identify the pathogens Shigella dysenteriae type 1, Escherichia coli (O157:H7), Campylobacter jejuni and Salmonella. A swab containing a stool specimen is inserted into the card. The card contains four microfluidic circuits which (1) capture and lyse the organism, (2) capture its nucleic acid, (3) amplify the nucleic acid, and (4) produce visual detection of the amplified prod-

produce shear stress. The field of biomechanics is concerned with the study of mechanical forces in living systems and the use of engineering design to create prosthetic devices and tools for rehabilitation. Understanding the complex interactions between the skeletal system, the muscular system and the nervous system required to produce coordinated movement is a challenging task. Experimental measurements during movement, such as the use of high speed cameras to track changes in the positions and orientations of body segments during motor tasks, coupled with surface electrodes to record the sequence and timing of muscle activity, have contributed greatly to our understanding of biomechanics. While these measurements can reveal data important to understand the kinematics and dynamics of body segment movement, they don’t explain how muscles work together at each instant during motor tasks. Over the past decade, large scale computational models have been developed to produce realistic simulations of movement that include a model of


Biomedical Engineering for Global Health

Profiles of translational innovation – Emil J. Freireich MD, Sc.D Emil J. Freireich, a founding father in the field of clinical cancer research, started his career intending to become a family physician. As he tells the story, “I grew up in the depression and my role model was my doctor because he was the only male that wore a tie, looked dignified and was educated. Everyone else was just digging holes and working for the WPA.” As fate would have it, every turning point in his career inadvertently led him away from family medicine toward the field of oncology. In 1955, Dr. Freireich (image courtesy of Emil J. Freireich MD, Sc.D., The University of Texas MD Anderson Cancer Center) became a Public Health Service Officer at the newly opened National Institutes of Health (NIH) in Bethesda, Maryland. As he explains, the physical layout of the NIH facility was revolutionary for its time. “It was designed to expedite the interaction between the basic sciences and the clinic . . . the [patient] wards were separated from the laboratories by only a service corridor. There is no hospital like it in the world.” Upon the suggestion of Dr. Gordon Zubrod, the Medical Director of the cancer institute, Dr. Freireich dedicated his efforts to finding a cure for leukemia, a cancer of the blood cells. In the early 1950s a diagnosis of leukemia was essentially a death sentence, with patients succumbing to massive hemorrhage and infection. Dr. Freireich recounts the reaction of Dr. Zubrod upon making rounds of the leukemia service.“He said to me, you know Freireich, your ward is a big mess. These children are really suffering. There’s blood everywhere! There’s blood on the pillows, on the sheets, the nurses are covered in blood, you’re covered in blood, it looks like a butcher shop. . . . You’re a hematologist Freireich, why don’t you do something about this bleeding?” Accepting the challenge, Dr. Freireich set out to discover the cause of the bleeding. Examining the clinical records of his patients over a period of several years, he identified a quantitative relationship between platelets, a component of the blood, and a propensity to hemorrhage. The lower a patient’s platelet count dropped, the greater the frequency and severity of hemorrhage. This groundbreaking work led him to hypothesize that transfusions of platelets could stop patients from hemorrhaging. Platelets were collected from donors using plasmapheresis, a technique whereby red blood cells are separated from the platelet rich plasma through centrifugation and returned to the donor. “We did a study where we matched up one donor with one child, [transfused] two units a week and could maintain them hemorrhage free for months.” Following implementation of a prophylactic platelet transfusion program, hemorrhage was controlled. “After that, we didn’t allow any blood on our ward. If there was blood on a pillow, we asked why didn’t this patient get platelets.” Despite the success, patients were still at risk for developing life threatening infections. Inspired by their work with platelets, Dr. Freireich and his colleagues questioned whether they could take the same approach with transfusions of leukocytes (white blood cells). Whereas platelets could be easily harvested from a donor’s blood, leukocytes proved more difficult for two reasons. First, normal healthy adults typically have very low numbers of leukocytes circulating in the blood at any given time. Second, leukocytes have density values very close to those of red blood cells, making it difficult to separate the cells from one another using simple centrifugation. Initially these problems were overcome by using donors with a condition called chronic myelogenous leukemia or CML. As Dr. Freireich explains, “We had patients that had chronic myelogenous leukemia and that disease is characterized by having neutrophil [leukocyte] counts two hundred times normal. That’s their disease. These neutrophils are not normal, they are leukemic neutrophils, but in the laboratory, functionally, they are about half as good as a granulocyte. . . . What if we collected granulocytes from donors with CML and transfused them into children?” Using this approach, Dr. Freireich demonstrated that infection in patients with acute leukemia could be controlled with daily transfusions of leukocytes. While this approach proved successful, it was not sustainable on a large scale. They needed a device that could collect blood from normal, healthy donors, selectively harvest the leukocytes, and return the red cells and plasma. Dr. Freireich went back to the lab. “I tried to use capillary flow and I had tubes all around my lab. Then I tried electromagnetic things, I tried different charges, I tried electropheresis, but all these techniques are cumbersome and slow. The only thing that was fast was the centrifuge.” While sitting in

The evolution of technology


his lab one day, Dr. Freireich was approached by George Judson, an IBM engineer whose son had come to the NIH for leukemia treatment. Judson wanted to help save his son’s life and joined Dr. Freireich in his quest to create a device that would separate blood according to its components. “One day he [Judson] appeared in my office with a pile of junk on a cart. He went to the IBM storeroom and found rejected pieces of plastic and screws and bolts and he actually built a centrifuge. He couldn’t test it, but it had all the ideas.” The initial design was relatively simple, relying on centrifugal force to separate the blood cells according to density and gravity to collect the components into separate containers. They tested the device in the lab using rejected blood, tweaking the device and adjusting the speed to achieve separation. Several iterations later, they were ready to test on a patient. “We wanted to do a CML patient because that would be easy. . . . We drew one unit of blood into the machine, separated it, and put nothing back, so it was perfectly safe and we showed it worked. The next step was to collect two units, separate it, and put one unit back.” In this manner, they worked their way up to a fully functioning device that came to be known as the continuous flow blood cell separator. (The continuous flow blood separators found in blood banks around the world are based on the original device shown in the photograph. Courtesy of NCI/G. Terry Sharrer, Ph.D., National Museum of History.) In 1965, the National Cancer Institute partnered with IBM to commercially produce the device to which Dr. Freireich holds the patent. It has since become a fixture in blood banks around the world. In just ten years, Dr. Freireich developed strategies to control hemorrhage and infection, revolutionizing the treatment of acute leukemia. For the first time in history, children diagnosed with leukemia had a fighting chance of survival. In light of Dr. Freireich’s commitment to educating the next generation of physicians and scientists, he offers a piece of advice. “Don’t think small . . . think about big things. It’s like hemorrhage or death or cancer or diabetes . . . the moon or space travel or energy. Young people have to think big, they have to tackle big problems. You have to be fearless.”

the skeleton, the muscle paths, musculo-tendon actuation, and excitation contraction coupling between the nervous system and the muscular system (Figure 7.21). As more detailed models are developed and validated, they have the potential to evaluate planned surgical procedures designed to correct gait abnormalities in patients who have experienced stroke or have cerebral palsy [18]. In addition to understanding motion, the field of biomechanics is also concerned with the way in which tissues respond to mechanical forces. It has been known for many years that tissues respond to mechanical forces. Over the past decade, it has become increasingly clear that cells themselves are exquisitely sensitive to mechanical forces and that changes in tissue structure that occur in response to mechanical forces begin with cellular changes. Mechanical forces can

initiate changes in gene expression which lead to protein synthesis, cell growth, death and differentiation. During normal growth and development, this mechanosensitivity is important to maintaining tissue homeostasis. For example, osteoblasts in bone are responsible for bone formation. They secrete proteins which make up the bone matrix. However, abnormal loading conditions can alter cell function and change the structure and composition of the extracellular matrix and produce pathologies such as osteoporosis, osteoarthritis, atherosclerosis, and fibrosis. The endothelial cells which line blood vessels secret matrix products as well as enzymes which break down structural proteins present in matrix. The endothelial cell response to abnormal forces in high blood pressure is an important component in the development of atherosclerosis, which can lead to heart attack [19].


Biomedical Engineering for Global Health constricted artery. A stent is a tubular scaffold made of metal which is inserted into the blood vessel to dilate the artery and restore flow; the stent must provide sufficient radial strength to hold the artery open. However, a stent can subject the artery to abnormally high stresses; these can lead to undesirable biological responses that cause restenosis and treatment failure. Biomechanical simulations can be used to investigate the effects of changing the geometry of a stent on the resulting arterial stress. Stents consist of concentric rings of sinusoid like curves connected by straight struts of varying length. Figure 7.22a shows the main parameters of a stent which can be varied: the spacing between struts in the stent, the radius of curvature of the small bends in each strut, and the height of these small bends. Figure 7.22b shows several generic stents that were designed by varying these parameters. Results of finite element models of stented arteries (Figure 7.22c) show that stent designs which incorporate large axial strut spacing, large radius of curvature, and high amplitudes will expose the smallest arterial segment to high stress and will still maintain sufficient blood flow [20]. Biomechanics bridges the fields of biology and mechanical engineering; work in this area provides opportunities to integrate experimental studies that operate across many scales from the molecular to cellular to tissue level. Progress in this area is dependent on close interaction between multi-scale modeling efforts and experimentation. The results of such work promises

Figure 7.21. Advances in modeling the movement of the human body have revolutionized the field of biomechanics. Used with permission from [18]. Reprinted, with permission, from the Annual c 2001 by Annual Review of Biomedical Engineering, Volume 3  Reviews

Atherosclerosis illustrates the interplay between biomechanics and mechanobiology. As we will see in Chapter 12, atherosclerotic blockages in the coronary arteries that supply blood to the heart can lead to heart attack. Mechanical interactions play an important role in the development of these blockages – whether they produce symptoms and whether they rupture and lead to a heart attack. An important treatment of atherosclerosis is the use of a stent to restore blood flow through a

to help understand basic physiologic processes, such as development, as well as important patho-physiologies.

Biomaterials and drug delivery The successful ability to implant biomedical devices and artificial tissues has revolutionized the treatment of many diseases, ranging from replacement heart valves to treat children with congenital heart disease, to artificial hip replacements to treat patients suffering from osteoarthritis, to surgically implantable polymer wafers which slowly release chemotherapy drugs to treat patients suffering from inoperable brain tumors. Each year, more than 200,000 pacemakers, 100,000 heart valves, one million orthopedic devices, and five million intraocular lenses are implanted into patients

The evolution of technology (a)


restore body function and come into contact with body fluids. In designing new biomaterials, it is important to understand both the chemical and mechanical requirements of the material. Because body chemistry is highly corrosive, and implanted materials may have to undergo many loading cycles per day, this is a particular challenge. In addition, the design of biomaterials must take


into account the interactions which will occur between the implanted materials and the surrounding tissue. Host responses, such as immune reactions, inflammation, wound healing, infection and tumor generation, can all occur in response to an implanted device. Understanding and controlling these reactions are crucial to clinical success. Original attempts to develop biomaterials focused on the design of passive, inert materials. However, trials in animal models and human subjects showed that the vast majority of materials elicit some type of cellular response in the host. Instead of attempting to design biomaterials that will simply act as static implants, the interdisciplinary field of bioma-


terials design has today evolved to focus on the design of materials that interact with tissue in a predictable manner, with the goal of creating a controlled cellular response between the artificial material and the living tissue surrounding it [21]. The challenges associated with developing effective biomaterials depend on whether the material is to be implanted permanently or temporarily and whether it is to be implanted within the body or on the body surface. Challenges are often easier to address for extracorporeal biomaterials, such as catheters, tubing, wound dressings, and dialysis membranes, which are placed temporarily outside the body, but come

Figure 7.22. (a) Stent parameters; (b) simulated stents; (c) hoop stress in stented arteries. Stress is lowest with design 2B3, which incorporates large strut length, large radius of curvature, and large amplitude [20].

worldwide; demand for implants of all kinds is growing at a rate of 5–15% each year [21]. The field of biomaterials engineering encompasses the design of any materials that are used to replace or

into contact with body fluids and tissues. Challenges increase for temporarily implanted biomaterials, which are designed to be placed inside the body and degrade over time, e.g. degradable sutures, implantable drug delivery systems, or scaffolds for cell or tissue transplants. Some of the greatest challenges arise with permanent implants, such as cardiovascular devices, orthopedic devices, dental devices, and sensory devices, which are designed to be placed within the body and must function effectively over a period of years to decades.


Biomedical Engineering for Global Health

A number of different types of synthetic and naturally occurring materials have been used to address the challenges associated with this broad range of clin-

ing polymer; together, this combination of materials provides both mechanical function as well as biological function. Early clinical results indicate that

ical applications. Ceramic materials, such as hydroxyapatite, calcium salts, and silicate ceramics, are often used to achieve hardness in implant surfaces such as those associated with joints or teeth. In addition, these materials can be easily bonded to bone surfaces to facilitate placement of an implant. They can also be used as the foundations of bone scaffolding materials in tissue engineering, where they can be manufactured to

this new implant substantially reduces the rate of restenosis in patients treated for coronary artery disease. The development of drug eluting stents illustrates the challenges that must be addressed in engineering new biomaterials. This design problem required an understanding of the biology of restenosis, development of drugs that target one or more pathways in the restenosis

degrade at controlled rates. Metals, such as titanium and stainless steel, are frequently used in implants that function in load bearing applications such as walking or chewing. Finally, the use of polymers provides the ability to design implants that have both flexibility and

process, as well as the development of a stent as a controlled delivery platform for release of drug. Restenosis occurs due to a complex cascade of events, which may include blood clot formation, inflammatory response, vascular smooth muscle cell proliferation, and synthe-

stability, and can be used in the design of articulat-

sis of extracellular matrix. Drugs which reduce the rate

ing surfaces which generate low friction. A number of synthetic biodegradable polymers are available (e.g. poly(glycolic acid), poly(ethylene glycol)), but biomaterials engineers have also derived polymers from natural

of restenosis have been identified and include compounds which suppress the patient’s immune response, reduce cellular proliferation, and reduce inflammatory response. When these drugs are given systemically, typ-

sources, such as modified polysaccharides, or modified proteins [21]. Recently, a new class of biomedical implants – the

ically they do not provide the desired effect. The challenge is that they are needed at high concentration only at the site of the stent. Biomaterials engineers focused

drug eluting stent – was developed to improve the treatment for cardiovascular disease. Most stents are made of stainless steel or nitinol (an alloy of nickel and titanium) [22]. Computational models have been developed

on developing stents which release the drug where it is needed. Stents coated with drug-loaded polymers can be used to provide release of drug at the site and time of injury with minimal systemic toxicity. Drug eluting

to predict the mechanical interactions between stent and artery wall, and we have seen how such models have been used to design stents which have a geometry that minimizes risk of clot formation and restenosis from the mechanical perspective [20]. However, this approach does not address the long-term biological response to the implanted stent. The biggest problem following placement of bare metal stents is restenosis of the vessel.

stents are coated with a drug-loaded polymer matrix which sustains drug release for up to four weeks following stent placement. The drug-releasing polymer coating is designed to be biologically inert and sterilizable and to be sufficiently flexible to follow changes in stent shape upon deployment [22]. The field of biomaterials and drug delivery integrates advances in materials science, polymer chemistry, phar-

Restenosis occurs as a result of trauma to the vessel wall induced by deployment of the stent; this trauma causes an aggressive healing response which overgrows the stent lumen [22]. Biomaterials engineers have developed special drug eluting stents which slowly release the drugs that

macology, and immunology. Implantable materials and drug delivery systems have the potential to impact a wide variety of disease ranging from improved therapy for advanced cancers to the development of systems to monitor and regulate blood glucose levels in diabetic patients. This field is characterized by multi-

inhibit this healing response. Drug eluting stents consist of a metal stent coated with a drug releas-

disciplinary studies bridging engineering, chemistry and biology.

The evolution of technology


Profiles of translational innovation – Robert Langer, Sc.D. Robert Langer, Sc.D. is a chemical engineer by training, but has taken a different road than most. In search of something more meaningful, Langer accepted a postdoctoral position with cancer researcher Judah Folkman, MD, at Children’s Hospital, in Boston. It was during this time that he began his revolutionary work in drug delivery. Today, he is a professor of chemical and biomedical engineering at the Massachusetts Institute of Technology where, together with physicians and researchers, he pushes the frontiers of biotechnology and materials science. Langer initially faced great skepticism and a general negative reaction from the science Courtessy of Stu Rosner Photograhpy. community upon proposing the use of polymers for the slow release of large molecules in a controlled manner. “I was very discouraged, but I just kept plugging. You write papers, you do talks, you do more experiments to convince the skeptics” [23]. However, three decades later Dr. Langer is distinguished among the most successful and renowned scientists in the biomedical engineering field. He has been one of the key pioneers in the field that paved the way for exciting new research in the areas of biomaterials and drug-delivery. “When I started doing this in 1974 there were almost no engineers working in medicine” [23]. When Langer began his work, researchers used off-the-shelf polymers for medical purposes. For example, the plastic used in women’s girdles was used in the first artificial heart because of their elastic properties and breast implants were initially filled with mattress stuffing. “This type of approach has often led to a number of problems. For example, when blood hits the surface of the artificial heart, a clot may form and the patient may suffer a stroke. We were interested in creating biomaterials that would have the right properties from the engineering, chemistry, and biological standpoint, and then synthesize them from first principles” [24]. It took a novel approach and Dr. Langer’s relentless effort to reach beyond this initial approach of biomaterials and ignite the field of biomedical engineering. Today, the fruits of Dr. Langer’s work can be seen globally, from nicotine patches to the stents currently implanted in cardiac patients. The chemotherapy wafers he developed with Dr. Henry Brem of Johns Hopkins have lead to the first new brain cancer treatment approved by the FDA in over twenty years. Artificial skin based on his research has been approved by the FDA and his work in artificial cartilage, bone, corneas blood vessels, spines and vocal cords is under development and appears promising. Even though Dr. Langer has nearly 550 issued or pending patents and is one of 13 Institute Professors at MIT (MIT’s highest honor) he continues to discover new treks to pursue in biomedical engineering. “Some of the newer things we’re looking at are solving problems with the delivery of genes, DNA, RNAi. Then there are remote control delivery, microchips, smart systems you can control- or maybe that you wouldn’t have to control” [25]. And as for the future of biomedical research, “One of the most important contributions we’ve made is to train the next generation. We’re bringing biomedical engineering to a different place and I think the contributions of those people will be ever greater as the years go on” [25].

Tissue engineering and regenerative medicine Over the past 50 years, advances in surgical techniques

mortality associated with end-stage organ failure, access to this procedure varies widely throughout the world

and the discovery of new immunosuppressive drugs have resulted in the development of organ transplantation as a tool to successfully treat end-stage organ failure of the kidney, liver, heart, lung, pancreas and intestine [26]. While organ transplantation has reduced

due to a combination of economic factors, availability of donor tissue, and access to specialist care. In the USA alone, more than 96,000 patients are waiting for transplant surgeries and 17 people die each day due to a shortage of donor organs [27, 28]. The field of tissue


Biomedical Engineering for Global Health naturally grow [29]. Consideration must be given to both the biomechanical and biochemical components of the microenvironment – a structure which provides the appropriate temporal and spatial sequence of signals dictating cell growth and differentiation is needed to yield a tissue that can survive and provide the appropriate biological function after implantation [29]. Scaffolds can be made of biological materials, synthetic materials, or hybrid biomaterials, so long as the scaffold provides cells with the right cues so that they even-

Figure 7.23. Cells from donor tissue are extracted, isolated, and placed on a porous biodegradeable scaffold that directs the growth of new cells. Once implanted into the body, the scaffold eventually disintegrates. Used with permission of Christine Schmidt, Ph.D., University of Texas at Austin.

tually synthesize and secrete their own matrix. New advances in tissue engineering are integrating advances in nano-structured composite materials with advances in drug delivery to provide targeted delivery of signaling molecules such as peptides, proteins and plasmid DNA and ensure repair of tissues in a timely manner.

engineering aims to integrate advances in engineering and life sciences to address this global need by devel-

Several clinical and commercial successes have been reported in the field of tissue engineering. A number of tissue engineered skin substitutes have been brought

oping living functional constructs that can be used to replace or regenerate damaged or diseased tissues in a less expensive and invasive manner.

to market in the last decade to treat burns, chronic ulcers, surgical wounds, and other dermatologic conditions (Figure 7.24) [31]. Apligraf is a two-layer tis-

The basic components used to engineer replacement tissues include: (1) cells to initiate development of new tissue (Figure 7.23a), (2) scaffolds to guide the three dimensional development of tissue (Figure 7.23b), and

sue engineered construct which mimics the structure of human skin. The bottom layer, the equivalent of the human dermis, is derived from neonatal foreskin fibroblasts in a contracted type I collagen matrix. The

(3) signals to coordinate cell growth and differentiation in space and time [29]. The goal of tissue engineering

top layer, the equivalent of the human epidermis, is generated from keratinocytes seeded onto the bottom

is to integrate these three components in a manner that promotes the development and growth of functional, three dimensional tissues. One strategy that has been successfully used to engineer tissues is to harvest cells from the intended recipient, expand these cells in culture outside the body,

layer. The cells in the construct do not survive long term after implantation, but instead encourage ingrowth of the patient’s own cells [31]. More recently, clinical trials have been carried out to transplant tissue engineered internal organs into patients. Anthony Atala at Wake Forest University

and then seed them onto a scaffold that drives formation of tissue until the cells can make their own supporting matrix [29]. Engineers have explored the use of scaffolds that are resorbed as new tissue grows as

School of Medicine led a team of engineers and clinicians who engineered a human bladder (Figure 7.25) and successfully transplanted it into patients. The team harvested about 1 million urothelial cells and muscle

well as permanent scaffolds. In any case, scaffolds must be biocompatible, and must provide a template for tissue growth in three dimensions. Since the phenotype of cells is very dependent upon their microenvironment, frequently engineers strive to design scaffolds which replicate the microenvironment in which cells would

progenitor cells from bladder biopsies of seven children with malfunctioning bladders as a result of spina bifida. Muscle cells were seeded onto dome-shaped, biodegradable molds of synthetic polymer and collagen, while bladder urothelial cells were seeded onto the interior surface. The constructs were grown in culture

The evolution of technology


Figure 7.24. Tissue engineered skin. c Jean Claude Revy, Phototake. Phototake 

has provided considerable excitement in the field of tissue engineering. Stem cells are undifferentiated cells that can proliferate, self-renew, and differentiate to one or more types of specialized cells when grown under appropriate conditions. Adult stem cells are rare undifferentiated cells found among the more common differentiated cells in a tissue that can renew themselves [29]. Adult stem cells are the basis for natural pathways of tissue maintenance and repair, and targeted activation of adult stem cells can turn on the body’s natural repair mechanisms. Embryonic stem cells are the most plastic Figure 7.25. Tissue engineered bladder. Courtesy of Wake Forest Institute of Regenerative Medicine.

for seven weeks, expanding the cells to about 1.5 billion in number, and then sewn to the patient’s own bladder. After implantation, the increased bladder capacity reduced the risk of long term kidney damage in this group of patients [32]. The design of engineered tissues must also take into account the further remodeling which will take place following implantation. Once tissue thickness exceeds sev-

cell source; they are totipotent – capable of differentiating to all cell lineages. Much research in the field of tissue engineering is now focused on how stem cells can be used as a source of cells in engineered tissues to reduce issues of immunogenicity and to increase the complexity of engineered tissues [29]. With current clinical successes and advances in tissue

eral hundred microns, it must be vascularized in order for cells to receive adequate oxgygenation [33]. The development of appropriate vascularization and innervation are significant challenges currently faced by the

engineering and regenerative medicine, the possibilities for the future include the development of an insulinsecreting, glucose-responsive bioartifical pancreas, the development of heart valves that can be implanted into children with congenital heart defects and can grow with the infant or child, as well as the repair or regeneration of the central nervous system [30]. The field of tissue engineering offers opportunities to combine

field. The recent development of techniques to isolate, culture, and differentiate embryonic and adult stem cells

advances in developmental biology, materials science, and engineering design in order to advance medical science.


Biomedical Engineering for Global Health

Systems biology and physiology Bioengineers seek to develop quantitative models of physiologic systems to help understand normal function and disease and to guide the design of therapeutic

At the cellular level, the focus is to develop models to understand how molecular pathways and networks relate to cellular functions such as metabolism, proliferation, differentiation and migration [34]. Developing

interventions. Organ level models of the cardiovascular system have elucidated the quantitative relationships between intracardiac pressure and tissue perfusion and are used to help physicians assess patients with heart disease or valve disorders and to design appropriate interventions. Mathematical models which describe coordinated nerve conduction, muscle contraction and musculoskeletal forces and motions have

computer models of cells (sometimes referred to as silicon cells) may enable one to test out potential drugs on the computer before they are tested in animals and used in clinical trials. This may have important global health implications; for example, Westerhoff and Bakker have developed a systems biology model of the parasite that causes African sleeping sickness, Trypanosoma brucei [37]. Through modeling, they discovered that a glucose

helped to understand normal gait as well as gait disorders [18]. Understanding the physiology of a whole organism is a complex task – organs are made up of tissues and cells and the goal of systems physiology is to describe

transporter is the best predicted drug target rather than the target which was previously under intensive study. Thus, systems biology may lead to more efficient drug discovery through the ability to quickly and efficiently carry out simulations to predict the effect of many differ-

the behavior of the system as a whole, starting with the

ent proposed drugs. A scientist may think that she has

component parts. Advances in the field of molecular biology have made tremendous progress toward understanding the molecular origins of physiology and disease. In traditional biology, the focus has been to study

developed an inhibitor that will block an enzyme pathway important in controlling disease. However, these pathways are highly connected into networks. Sometimes the effect of blocking one pathway results in unan-

individual genes or proteins one at a time; however, in systems biology, the focus is to investigate the behavior and relationships of all the elements in the system and

ticipated effects – systems biology provides a way to anticipate these interactions without having to do extensive experimentation [36, 37].

to describe the interactions among them as part of one functioning system [34]. It is now possible to rapidly assess changes in gene expression, protein expression, and protein–protein

Advances in the field of systems biology and computational bioengineering over time will lead to the ability to model increasingly complex physiological systems. Some success has been achieved to develop integrative

interactions in cells and tissues. These new experimental techniques have generated extremely large and complex datasets, driving the need for models to pull them together in a way that helps us understand biology at a higher level, as a complex collection of networks and pathways. This approach is important because the causes of many common diseases are multi-factorial in nature –

systems physiology models of the cardiovascular system. These cell models have been integrated into large scale, anatomically detailed models of electrical conduction to investigate the molecular basis of life threatening arrhythmias [34]. The field is characterized by close interplay between experts in the areas of genetics, molecular and cell biology, statistics, and computer science.

they are not caused by changes in a single gene or protein, and are often not treated effectively with a single drug [35]. However, using systems biology and physiology to construct models of complex biological systems and diseases may aid in understanding complex pathophysiology and may guide the search for multi-target approaches to drug therapy [34].

Molecular and cellular engineering Molecular and cellular engineering uses engineering principles to understand and construct cellular and molecular systems with useful properties. At the molecular level, proteins can be engineered to modify the communication of cells with their environment; this

The evolution of technology


approach can form the basis for rational design of targeted drug therapies to treat cancer. At the cellular level, metabolic engineering can be used to design cellular factories which manufacture pharmaceuticals or scaffolds for use in tissue engineering applications, or to be used as cellular biosensors to monitor the environment for toxic chemicals. At the molecular level, much research in this field is focused on discovering the design principles that govern the behavior of macromolecular complexes and interacting networks of proteins found within cellular organelles, molecular motors and biological membranes. Complexes of biological macromolecules form the basis of many cellular processes, including signaling, motility, metabolism, and biomolecular transport. Bioengineers are developing new computational and experimental methodologies that provide unique insights into how biological macromolecules self-assemble, interact, and function collectively. Understanding the function of these supramolecular assemblies is essential to understanding complex biomolecular processes, and will create new avenues to predict, control, and manipulate biomolecular machinery. As an example, cells contain a wide variety of biological nanomotors (Figure 7.26): flagellar motors propel bacteria; motor proteins, such as myosin, are responsible for muscle contraction; RNA based motors enable packaging of viral nucleic acids when viruses reproduce within host cells. Kinesin is a motor protein important in organelle transport and mitosis [38]. These motors are remarkable for their efficient conversion of chemical energy into mechanical work – they operate at greater than 50% efficiency, double that of the average engine used in cars. Combining tools of molecular biology with single molecule imaging and force measurements is providing a clearer picture of how these molecular motors operate within cells; this knowledge can then be translated to harness nanomotors to power nanodevices in analytical biosensors or molecular assembly platforms. Biological nanomotors may provide a solution to the difficulties of moving biological fluids through nanofluidic devices. Moving solutions through nanodevices is particularly challenging because the ratio of surface area

Figure 7.26. Examples of biological nanomotors: kinesins move along microtubules to move cargo toward the periphery of a cell. Dynein moves in the opposite direction, to transport cargo to the center of the cell. Mysoin motors move along actin filaments. F1-ATPase is a rotary motor [38].

to volume is high in these devices, dramatically increasing the effects of friction. An alternative approach is to bind the analyte of interest to a molecular shuttle (Figure 7.27) powered by a molecular motor; the motor can then be used to move the molecule of interest while leaving the bulk of the solution at rest [38]. Similarly, there is great interest in using molecular motors to direct and control macromolecular assembly. Exploiting the understanding of biological motors, efforts are underway to design networks of nanoscale conveyor belts which transport molecules and control their encounters with reaction partners in order to yield prescribed target products. At the cellular level, engineers use quantitative tools to understand and manipulate the network of metabolic reactions within the cell. Using the techniques of molecular biology, it is now possible to modify specific enzyme controlled reactions within the metabolic network of a cell. Cellular engineering refers to the improvement of cell properties through modification of specific biochemical reactions. An important goal of cellular engineering is to develop cell systems with desired properties, such


Biomedical Engineering for Global Health

Example metabolic network model for Escherichia coli. The model incorporated data on 436 metabolic intermediates undergoing 720 possible enzyme-catalyzed reactions. Circles show abbreviated names of metabolic intermediates, and arrows represent enzymes. Heavy lines indicate links with high metabolic fluxes. From [39]. Figure copyright 2000, National Academy of Sciences USA.

Figure 7.27. A proposed molecular shuttle system [38].

as improved ability to synthesize natural products of interest, the ability to produce products that are new to the host cell, or improved ability to function in extreme environments such as hypoxic conditions [40]. The use of recombinant techniques, combined with cellular engineering, has improved the ability to engineer cells to produce protein pharmaceuticals such as insulin. There are currently more than 200 FDA approved peptide and protein pharmaceuticals. Natural sources of these compounds are often rare and expensive. Today, most of these are produced using recombinant methods in bacteria, yeast, animal cells, or plants [41]. The field of cellular engineering has also made contributions to understand and manipulate the interactions between a cell and its environment. The bi-directional communication between cells and their environment is referred to as cell signaling. Cell signaling controls many complex biological processes, such as development, tissue function, immune response, and wound healing.

In communicating with their environment, cells need to receive environmental signals, react to these signals by translating them into appropriate intracellular responses, and, if necessary, by sending an extracellular message back to the environment. Many diseases result from a breakdown in this communication: auto-immune diseases result from the failure of cells to correctly read signals (self vs foreign); in cancer, genetic mutations can hardwire a cell’s signaling machinery in a pro-growth state, leading to uncontrolled growth even in absence of external signals to stimulate growth. Thus, understanding cellular signaling provides an opportunity to

The evolution of technology gain insight into a wide variety of disease processes, and the molecules important in cell signaling provide good targets for disease therapy [42]. One of the most studied cell signaling systems is the family of tyrosine kinase receptors, which includes the epidermal growth factor receptor (EGFR). These receptors mediate the signaling network that transmit extracellular signals into the cell, controlling cellular differentiation and proliferation [43]. Normally their activity is tightly regulated. Models of tyrosine kinase receptors have been instrumental in understanding both basic cell biology as well as important clinical features, such as the propensity for cancer cells to metastasize [42]. Studies of cell signaling and the tyrosine kinase receptors have led to new drugs which inhibit these receptors. Herceptin, Gleevec and Iressa are the first examples of targeted therapeutics for tyrosine kinase receptors and are used to treat advanced breast cancer, chronic myelogenous leukemia and gastro-intestinal stromal tumors, and lung cancer, respectively [43]. The field of cellular and molecular engineering is contributing both to our knowledge of how biological systems are organized and interact as well as to the development of new therapeutic molecules and ways to produce them efficiently and inexpensively. This field is characterized by collaboration between biochemists, biologists and engineers [44]; interaction between combinatorial experimental approaches and large scale computational modeling is particularly important to advances in this field.

Bioengineering and biotechnology to improve health in developing countries Many of the technologies that we have just seen are available primarily in developed countries (recall that MRI systems cost millions of dollars and most bioengineering research requires expensive computational, instrumental, and material infrastructure). A recent panel of 28 scientific experts from around the world who are well acquainted with health problems of developing countries was asked “What do you think are the major biotechnologies that can help improve health in developing countries in the next 5–10 years?” [45]. The top ten technologies are profiled in Table 7.3. As we have


Table 7.3. Top ten biotechnologies for improving health in developing countries [45]. Rank



Modified molecular technologies for affordable, simple diagnosis of infectious diseases


Recombinant technologies to develop vaccines against infectious diseases


Technologies for more efficient drug and vaccine delivery systems


Technologies for environmental improvement (sanitation, clean water, bioremediation)


Sequencing pathogen genomes to understand their biology and to identify new anti-microbials


Female controlled protection against sexually transmitted disease, both with and without contraceptive effect


Bioinformatics to identify drug targets and to examine pathogen–host interactions


Genetically modified crops with increased nutrients to counter specific deficiencies


Recombinant technology to make therapeutic products (for example insulin, interferons) more affordable


Combinatorial chemistry for drug discovery

seen in this chapter, the field of bioengineering plays an important role in the development of many of these technologies, including the development of inexpensive tools for point-of-care detection of infectious disease, methods to more efficiently deliver drugs and vaccines, computational and combinatorial methods to develop new therapeutic products, and the use of recombinant techniques to produce drugs in a more cost effective manner. Currently, most research efforts in the field of bioengineering and biotechnology are focused on health challenges faced by developed countries, and new devices are designed to meet the constraints present in laboratory and healthcare facilities in the developed world. It has been estimated that 90% of the health research dollars are spent on the health problems of 10% of the world’s population [45]. Between 1975 and 1997, only 13 new chemical entities were developed for the treatment of tropical diseases [46]. Barriers which limit the development and dissemination of new technologies


Biomedical Engineering for Global Health

for the developing world include low profit margins in developing countries, lack of infrastructure, and regulatory constraints [47].

committee of experts advising the G8 recommended initially using guaranteed markets as a way to develop a vaccine for pneumococcal disease which kills more

The importance of market forces increasingly plays a role in determining which potential new products receive private investment; the cost of bringing a new medicine to market in the USA has been estimated to be as high as $0.8–$1.7 billion [48]. This is a major disincentive to investment in drugs for rare diseases or those that predominantly affect the developing world. Recently, a number of new ideas have been proposed to

than 1.5 million people every year, many under the age of five. Guaranteed markets as a tool to encourage the development of a vaccine to prevent malaria are also of great interest [52]. In the next chapters, we will examine in detail the development of several classes of technologies designed to improve health. We will focus on new tools to treat, detect and prevent the leading causes of death through-

address the failure of market forces to lead to technologies to address the health needs of developing countries. The USA has made a substantial increase in its support for biomedical research; from 1998 to 2005, the budget of the National Institutes of Health (NIH)

out the world: infectious disease, heart disease and cancer. In Chapter 8, we will examine technologies to prevent infectious diseases, beginning with an overview of how our immune system protects us against disease, and then considering the steps involved in designing

doubled to nearly $28 billion annually; and as of 2004,

new vaccines to protect against disease. In Chapter 10,

the USA spent 0.25% of its GDP per year on health related research, more than double the average of other developed countries [49, 50]. However, the focus of the NIH is largely to address health priorities of importance

we will develop an understanding of the biology of cancer, and will examine the development of new technologies to detect cancers at a stage when they are still treatable, as well as technologies to prevent the

in the USA. Some emerging economies, notably those of South Korea, China, and India, have benefited from strong increases in public investment in scientific and

development of cancer. Finally, in Chapter 12, we will examine cardiovascular physiology and pathophysiology and we will consider the engineering of technologies

health-related research; for example, in South Korea, the number of health biotechnology related publications by South Korean researchers increased by tenfold from 1992 to 2002 [51]. A number of private organizations,

designed to treat heart disease as well as approaches to prevent heart disease. Along the way, we will find we need several additional tools to facilitate the translation of new technologies. In Chapter 9, we will consider

including the Carter Center, and the Bill and Melinda Gates Foundation have made substantial contributions to research and development focused on meeting the health needs of developing countries. Another approach to stimulate private investment in diseases that affect developing countries is for governments to guarantee markets for new technologies. In 2006, the G8 countries considered a plan to provide

the ethical guidelines which have been developed to

a guaranteed market for new vaccines that meet predetermined safety and efficacy standards [52]. As a way to incentivize private investment in developing new vaccines, governments have agreed to provide subsidies ranging from $800 million to $6 billion depending on the disease to purchase the resulting vaccine. Once the initial subsidies have been spent, pharmaceutical companies would be required to provide vaccine to developing world customers at sharply discounted prices. A

Bioengineering and Global Health Project Project task 3: Evaluate current policy designed to develop or implement solutions to the problem What investments is the health system in the region you have selected making to develop or implement new solutions? Are there other efforts from the private or public sectors to develop new solutions? Are there investments in basic or applied research? Are large clinical trials underway to test new solutions? What are the limitations of these approaches? Write a one page summary of current health policy regarding the health problem/region you have selected.

The evolution of technology ensure that the rights of human subjects participating in clinical trials of new technology are adequately protected. In Chapter 11, we will learn how to assess the cost effectiveness of new interventions. In Chapter 13, we will learn how to design clinical trials and to choose a sample size to achieve statistically meaningful results.

Homework 1. Read the following abstract from an article recently published in Nature. Briefly explain how the steps the authors took correspond to the steps of the scientific method. All five steps are represented here. Letter: An unexpected cooling effect in Saturn’s upper atmosphere C. G. A. Smith, A. D. Aylward, G. H. Millward, S. Miller and L. E. Moore n7126/abs/nature05518.html. Reprinted with permission from Macmillan Publishers Ltd: Nature, Smith et al. An unexpected cooling effect in Saturn’s upper atmosphere, 445(7126): 399–401. Copyright 2007. The upper atmospheres of the four Solar System giant planets exhibit high temperatures that cannot be explained by the absorption of sunlight. In the case of Saturn the temperatures predicted by models of solar heating are 200 K, compared to temperatures of 400 K observed independently in the polar regions and at 30◦ latitude. This unexplained “energy crisis” represents a major gap in our understanding of these planets’ atmospheres. An important candidate for the source of the missing energy is the magnetosphere, which injects energy mostly in the polar regions of the planet.


large – source of polar energy, or, more probably, some other process heats low latitudes directly. 2. Compare and contrast the first steps in the engineering and scientific methods. Why are these differences important? 3. Directions: After each description of health news just released in 2004, identify whether you think it would be better labeled as science or engineering, then briefly describe what characteristics of the example support your choice. (A) Laboratory rat gene sequencing completed; humans share one-fourth of genes with rat, mouse A large team of researchers, including a computer scientist at Washington University in St. Louis, has effectively completed the genome sequence of the common laboratory brown rat, Rattus norvegicus. This will make the third mammal to be sequenced, following the human and mouse. 3222.html (B) Chemists seek light-activated glue for vascular repair Surgeons battle time and the body’s defenses as they stitch together veins and arteries, whether after an injury or in the course of such treatments as transplants or bypasses. Loss of blood before a site is closed and too much clotting soon after challenge medical care. Virginia Tech researchers are creating biocompatible adhesives for use with vascular tissue that will speed the process of mending tissue.

This polar energy input is believed to be sufficient to explain the observed temperatures, provided that it is efficiently redistributed globally by winds, a process that is not well understood. Here we show, using a numerical model, that the net effect of the winds driven by the polar energy inputs is not to heat but to cool the low-latitude thermosphere. This surprising result allows us to rule out known polar

(C) New biomaterial may replace arteries, knee cartilage A unique biomaterial developed by researchers at the Georgia Institute of Technology could be available in as few as five years for patients needing artery or knee cartilage replacement. It may also be used to speed repair of damaged nerves in patients with

energy inputs as the solution to the energy crisis at Saturn. There is either an unknown – and

spinal cord injuries and as the basis for an implantable drug delivery system.


Biomedical Engineering for Global Health newsrelease/BIOMAT.html (D) New insight on cell growth could lead to method for stopping cancer West Lafayette, Ind. – Halting the development of certain pancreatic, ovarian, colon and lung cancers may be possible with therapy based on recent Purdue University research. By investigating a single molecule that influences cell growth, a research group in the Purdue Cancer Center, has gained new insight into the chain of events that make some cancer cells divide uncontrollably – insight that may eventually lead to a way to break that chain, stopping cancer in its tracks. 2004/040328.Henriksen.ras.html 4. Identify a researcher involved in developing new health technologies at your institution. Arrange to interview them in person. Write a one page profile of this individual. Your profile should include at least the following. r A summary of the researcher’s educational background. r A description of current technologies they are developing and the potential of these technologies to improve health. r Your assessment of how or where their research and development efforts fit along the spectrum of translational research.

References [1] Tredgold T. ICE Council Minutes. ICE Meeting of Council. 1827–1835 December 29;3. [2] National Science Foundation. Integrative Graduate Education and Research Traineeship. 2007 [cited 2007 June 1]; Available from: [3] National Institutes of Health. NIH Roadmap for Medical Research. 2007 May 25 [cited 2007 June 1]; Available from: [4] Hopkin K. Biography of Huda Y. Zoghbi, M.D. Chevy Chase, M.D.: Howard Hughes Medical Institute; 2003. [5] Heart–Lung Machine. Arlington, VA: The Whitaker Foundation; 2006. [6] Stephenson LW. History of cardiac surgery. In: Edmonds LH, Cohn LH, eds. Cardiac Surgery in the Adult. New York: McGraw-Hill; 2003.

[7] Kasper D, Braunwald E, Fauci A, Longo D, Hauser S, Jameson JL. Harrison’s Principles of Internal Medicine. 16th edn. New York: McGraw-Hill; 2005. [8] American Heart Association. Heart Disease and Stroke Statistics-2007 Update. Dallas, TX: American Heart Association; 2007. [9] History of Medical Diagnosis and Diagnostic Imaging. 2006 August [cited 2006 September 16]; Available from: [10] Lebedev MA, Nicolelis MA. Brain–machine interfaces: past, present and future. Trends in Neurosciences. 2006 Sep; 29(9): 536–46. [11] Chapin JK, Moxon KA, Markowitz RS, Nicolelis MA. Real-time control of a robot arm using simultaneously recorded neurons in the motor cortex. Nature Neuroscience. 1999 Jul; 2(7): 664–70. [12] Fetz EE. Real-time control of a robotic arm by neuronal ensembles. Nature Neuroscience. 1999 Jul; 2(7): 583–4. [13] Brain-Computer Interfaces Come Home: National Institute of Biomedical Imaging and Bioengineering; 2006 November 28. [14] Bonetta L. Flow cytometry smaller and better. Nature Methods. 2005; 2(10): 785–95. [15] Rodriguez WR, Christodoulides N, Floriano PN, Graham S, Mohanty S, Dixon M, et al. A microchip CD4 counting method for HIV monitoring in resource-poor settings. PLoS Medicine. 2005 Jul; 2(7): e182. [16] Yager P, Edwards T, Fu E, Helton K, Nelson K, Tam MR, et al. Microfluidic diagnostic technologies for global public health. Nature. 2006 Jul 27; 442(7101): 412–18. [17] PATH. Rapid Diagnostic Test Technologies: Lateral-flow. [cited 2007 April 10]; Available from: [18] Pandy MG. Computer modeling and simulation of human movement. Annual Review of Biomedical Engineering. 2001; 3: 245–73. [19] Wang JH, Thampatty BP. An introductory review of cell mechanobiology. Biomechanics and Modeling in Mechanobiology. 2006 Mar; 5(1): 1–16. [20] Bedoya J, Meyer CA, Timmins LH, Moreno MR, Moore JE. Effects of stent design parameters on normal artery wall mechanics. Journal of Biomechanical Engineering. 2006 Oct; 128(5): 757–65. [21] Capturing the Full Power of Biomaterials for Military Medicine: Report of a Workshop. Washington, D.C.: National Academy of Sciences; 2004. [22] Burt HM, Hunter WL. Drug-eluting stents: a multidisciplinary success story. Advanced Drug Delivery Reviews. 2006 Jun 3; 58(3): 350–7. [23] Saulnier B. The Big Picture. Cornell Alumni Magazine. 2004 Sep/Oct; 107(2).

The evolution of technology [24] Langer RS. An interview with a distinguished pharmaceutical scientist: Robert S. Langer. Pharmaceutical Research. 1999 Apr; 16(4): 475–7. [25] Langer R. Robert Langer, ScD – Engineering medicine. Interview by M. J. Friedrich. Jama. 2005 Oct 5; 294(13): 1609–10. [26] Timeline of Key Events in U.S. Transplantation and UNOS History. 2007 [cited 2007 April 12]; Available from: [27] U.S. Transplantation Data. 2007 [cited 2007 April 12]; Available from: default.asp?displayType=usData [28] Help Save a Life. 2007 [cited 2007 April 12]; Available from: [29] Polak JM, Bishop AE. Stem cells and tissue engineering: past, present, and future. Annals of the New York Academy of Sciences. 2006 Apr; 1068: 352–66. [30] Nerem RM. Tissue engineering: the hope, the hype, and the future. Tissue Engineering. 2006 May; 12(5): 1143–50. [31] Ehrenreich M, Ruszczak Z. Update on tissue-engineered biological dressings. Tissue Engineering. 2006 Sep; 12(9): 2407–24. [32] Atala A, Bauer SB, Soker S, Yoo JJ, Retik AB. Tissue-engineered autologous bladders for patients needing cystoplasty. The Lancet. 2006 Apr 15; 367(9518): 1241–6. [33] Aschheim K. Profile: Anthony Atala. Nature Biotechnology. 2006 Nov; 24(11): 1311. [34] Ideker T, Winslow LR, Lauffenburger AD. Bioengineering and systems biology. Annals of Biomedical Engineering. 2006 Feb; 34(2): 257–64. [35] Institute for Systems Biology. Health Care in the 21st Century: Predictive, Preventive and Personalized. 2006 [cited 2006 November 13]; Available from: Intro to ISB and Systems Biology [36] Westerhoff HV, Palsson BO. The evolution of molecular biology into systems biology. Nature Biotechnology. 2004 Oct; 22(10): 1249–52. [37] Henry CM. Systems biology. Chemical and Engineering News. 2003 May 19: 45–55. [38] Hess H, Bachand GD, Vogel V. Powering nanodevices with biomolecular motors. Chemistry (Weinheim an der Bergstrasse, Germany). 2004 May 3; 10(9): 2110–16. [39] Edwards JS, Palsson BO. The Escherichia coli MG1655 in silico metabolic genotype: its definition, characteristics, and capabilities. Proceedings of the





[44] [45]









National Academy of Sciences of the United States of America. 2000 May 9; 97(10): 5528–33. Stephanopoulos G. Metabolic fluxes and metabolic engineering. Metabolic Engineering. 1999 Jan; 1(1): 1–11. Wishart D. Production of Protein Pharmaceuticals (Part 1) [PowerPoint Slides]: University of Alberta; 2005. Asthagiri AR, Lauffenburger DA. Bioengineering models of cell signaling. Annual Review of Biomedical Engineering. 2000; 2: 31–53. Bennasroune A, Gardin A, Aunis D, Cremel G, Hubert P. Tyrosine kinase receptors as attractive targets of cancer therapy. Critical Reviews in Oncology/ Hematology. 2004 Apr; 50(1): 23–38. Maynard J, Georgiou G. Antibody engineering. Annual Review of Biomedical Engineering. 2000; 2: 339–76. Daar AS, Thorsteinsdottir H, Martin DK, Smith AC, Nast S, Singer PA. Top ten biotechnologies for improving health in developing countries. Nature Genetics. 2002 Oct; 32(2): 229–32. Pecoul B, Chirac P, Trouiller P, Pinel J. Access to essential drugs in poor countries: a lost battle? Jama. 1999 Jan 27; 281(4): 361–7. Free MJ. Achieving appropriate design and widespread use of health care technologies in the developing world. Overcoming obstacles that impede the adaptation and diffusion of priority technologies for primary health care. International Journal of Gynaecology and Obstetrics: the Official Organ of the International Federation of Gynaecology and Obstetrics. 2004 Jun; 85 Suppl. 1: S3–13. Innovation or Stagnation? Rockville, MD: US Department of Health and Human Services, Food and Drug Administration; 2004 March. Koizumi K. National Institutes of Health in the FY 2007 Budget. In: Intersociety Working Group, ed. AAAS Report XXXI Research and Development FY 2007. Annapolis Junction, MD: AAAS 2006. A.8. Health Related R&D. OECD Science, Technology and Industry Scoreboard 2005 – Towards a knowledge-based economy OECD Publishing 2005. Wong J, Quach U, Thorsteinsdottir H, Singer PA, Daar AS. South Korean biotechnology – a rising industrial and scientific powerhouse. Nature Biotechnology. 2004 Dec; 22 Suppl.: DC42–7. Phillips M. Global Vaccine Initiative Hits Snag. Wall Street Journal. 2006 July 7.

8 Prevention of infectious disease

In Chapter 7, we examined the process of designing new technologies and the interdisciplinary translational research efforts needed to advance technologies from the laboratory to clinical practice. In the rest of this book, we will explore in detail the development of several types of new technologies which draw on advances in the different sub-disciplines of bioengineering. We begin by focusing on the development of vaccines to prevent infectious disease. We will see that scientific knowledge, such as an understanding of both the organisms that cause disease and the protective mechanisms of the immune system, is critical to enable the engineering of preventive vaccines. We will examine the development of vaccines from idea to product. We have seen that infectious diseases are responsible for a large fraction of deaths, particularly in the developing world. In high-mortality developing countries, infectious disease is responsible for nearly half of all deaths, and the childhood cluster diseases (pertussis, poliomyelitis, diphtheria, measles, and tetanus) kill just over half a million children under the age of five each year [2]. In the developed world, the use of vaccines has dramatically reduced the incidence of infectious disease. As we explore how vaccines work, we will trace the tremendous improvements in world health that have resulted from mass childhood immunization and examine the global obstacles that remain. There are many

Figure 8.1. Vaccines play a critical role in the prevention of infectious diseases.

Prevention of infectious disease

Progress to fight infectious diseases in the USA “The gasping breath and distinctive sounds of whooping cough; the iron lungs and braces designed for children paralyzed by polio; and the devastating birth defects caused by rubella: To most Americans, these infectious scourges simultaneously inspire dread and represent obscure maladies of years past. Yet a little more than a century ago, the US infant mortality rate was a staggering 20%, and the childhood mortality rate before age 5 was another disconcerting 20% . . . Fortunately, many of these devastating diseases have been contained, especially in industrialized nations, because of the development and widespread distribution of safe, effective and affordable vaccines”[1].

diseases for which no vaccine is available. We conclude this chapter by considering the scientific and engineer-


of high dose chemotherapy for breast cancer. A side effect of that treatment was that it destroyed cells in the patient’s bone marrow, leaving patients vulnerable to infection. Stem cells in the bone marrow are responsible for generating the cellular components of blood: the oxygen carrying red blood cells, the platelets which aid in blood clotting and the infection–fighting white blood cells. There are five main types of white blood cells: eosinophils, neutrophils, basophils, lymphocytes and monocytes (Figure 8.2). Eosinophils are important in fighting infections due to parasites, such as malaria. Neutrophils are important in fighting infections caused by bacteria, such as tuberculosis. Monocytes leave the bloodstream and mature into macrophages, which are also important in fighting bacterial infections. As we will see, lymphocytes are important in the fight against both bacterial and viral infections. The function of basophils is poorly understood, but they are believed to be important in allergic reactions. To understand how white blood cells work as part of the immune system to fight infection, we next examine how infectious agents, such as bacteria and viruses, cause disease [3].

ing challenges associated with developing new vaccines to prevent HIV infection.

How infectious agents cause disease The immune system Vaccines manipulate the immune system of the recipient; thus to understand how vaccines work, we must first understand how the immune system prevents and fights infections. In Chapter 1, we examined the use



Bacteria and viruses cause disease in very different ways. Bacteria consist of cells with a cell membrane and, unlike human cells, usually also have a cell wall (Figure 8.3). Bacteria can survive outside of a host and are capable of reproduction without a host.


Figure 8.2. Three of the main types of white blood cells are (a) neutrophils, (b) lymphocytes, and (c) macrophages.


Biomedical Engineering for Global Health



(b) (b)

Figure 8.3. (a) The structure of a bacterial cell and (b) an electron micrograph of bacterial cells. CDC/Dr. Ray Butler, Janice Carr.

Bacteria can be killed or inhibited by antibiotics, which frequently destroy the bacterial cell wall. Viruses consist of a nucleic acid core surrounded by a protein envelope (Figure 8.4). Unlike bacteria, viruses must use the intracellular machinery of their host to reproduce, and they cannot be killed with antibiotics. There are more than 50 different viruses that can infect humans [4]. Whether virus or bacteria, there are three basic problems each pathogen must solve: (1) how to reproduce inside a human host, (2) how to spread from one person to another, and (3) how to evade the immune system. Because of their fundamental structural and functional differences, bacteria and viruses solve these problems and cause disease in very different ways. Bac-

Figure 8.4. (a) The structure of a virus and (b) an electron micrograph of a virus. CDC/ Dr. F. A. Murphy CDC.

teria invade a host, and then begin to reproduce. As they grow and reproduce, they produce toxins which disturb the function of normal cells. Viruses actually invade the cells of their host (Figure 8.5). They typically accomplish this invasion by binding to receptors on the membrane of the host cell, which then transport the virus into the cell in a process known as endocytosis. Once inside, the virus takes over the cell, using viral nucleic acid and

Prevention of infectious disease


Figure 8.5. The life cycle of a virus. Reprinted with permission from Macmillon Publishers, Ltd: Nature. Brett D. Lindenback: Charles M. Rice. Unravelling hepatitis C virus replication from genome to function. Nature Publishing Group. 2005. 436(7053): 933–938.

host cell resources to make new viral nucleic acid and proteins. As the virus directs the synthesis of new viral

gasp is accompanied by a high pitched sound (whoop). Pertussis is caused by the bacterium Bordetella pertus-

particles, more virus is released from the host cell. The new virus is disseminated when the virus either causes the host cell to lyse (break apart) or the newly formed viral particles bud from host cell surface [3].

sis. How does the Bordetella pertussis bacterium cause these symptoms? The respiratory tract is lined with

Bacterial disease Let’s examine a once common bacterial infection – pertussis, also known as whooping cough – to see how bacteria typically cause disease. Pertussis is a highly infectious respiratory illness which is transmitted through contact with respiratory droplets from an infected person. Typically seven to ten days after exposure, patients develop cold-like symptoms, such as a runny nose, sneezing, low grade fever, and a mild cough. The cough gradually worsens, and after one to two weeks patients experience violent bursts of coughing. At the end of a coughing episode, a long

ciliated epithelial cells; ordinarily these cilia play an important role in clearing the respiratory tract of mucus and secretions. The Bordetella pertussis bacterium binds to receptors on the surface of cells lining the respiratory tract (Figure 8.6); as the bacteria grow, they produce toxins which paralyze the respiratory cilia and interfere with the patient’s ability to clear respiratory secretions. The violent coughing fits associated with pertussis are an attempt to clear secretions which build up due to impaired ciliary function. Ultimately, the immune system usually clears the bacterial invasion, and patients recover within several weeks. However, infants and young children are at particular risk of developing a secondary bacterial pneumonia which can lead to death [1].


Biomedical Engineering for Global Health

[5] At the end of a pertussis coughing episode, a long gasp is accompanied by a high pitched sound (whoop). You can listen to a typical “whooping” cough associated with pertussis at: http://www. Before the availability of the pertussis vaccine in the 1940s, more than 200,000 cases of whooping cough were reported each year in the USA. The disease is still an important cause of mortality in developing countries; in 2001, pertussis was estimated to cause more than 285,000 deaths in children in developing countries. Figure courtesy of CDC vaccines/pubs/pinkbook/downloads/pert-508.pdf

Figure 8.7. Influenza is spread from person to person through coughs or sneezing. Courtesy of Andrew Dandhazy, Rochester Institute of Technology.

Viral disease In contrast, viral pathogens have developed different ways to reproduce in a host, spread from one person to another, and evade the immune system. How does a common virus like influenza solve these problems? In order to reproduce, the influenza virus must get inside the human cell to use the cell’s biosynthetic machinery. Influenza virus accomplishes this by binding to receptors on the host cell surface and then inducing receptor mediated endocytosis. Once it has been endocytosed, the influenza virus is trapped in a vesicle made of the cell membrane called an endosome. The virus acts to slowly reduce the pH in the endosome, creating a hole in the membrane through which the virus releases its single stranded RNA and polymerase proteins. These RNA segments and polymerase proteins enter the nucleus of the infected cell and direct the cell to begin making many copies of the viral RNA and viral coat proteins. These new viral particles then exit the nucleus and bud from

Figure 8.6. Bordetella pertussis bacteria bind to the cilia of cells lining the respiratory tract. Reprinted from Respiratory Medicine, Vol. 94, Soane, M. C. et al., Interaction of Bordetella pertussis with c 2000, with human respiratory mucosa in vitro, pp. 791–799,  permission from Elsevier.

the cell. During this reproduction, the viral polymerase proteins don’t proofread reproduction, and as a result nearly every virus produced in an influenza-infected cell is a mutant, differing slightly from the original infecting virus [4]. How does the influenza virus spread from one person to another? Generally, this happens when an infected person sneezes or coughs (Figure 8.7), and micro-droplets containing viral particles are inhaled by

Prevention of infectious disease


called reverse transcriptase surrounded by a protein core (Figure 8.8). The entire virus is surrounded with a lipid membrane; this membrane is studded with special proteins called gp120 envelope proteins that enable it to bind to the surface of host cells. In particular, this envelope protein is recognized by receptors on the surface of a special type of lymphocyte [3]. Figure 8.9 illustrates what happens when the HIV virus binds to the surface of a host lymphocyte. The membrane of the viral particle fuses with the host cell membrane, allowing the contents of the viral particle to enter the cytoplasm of the lymphocyte. The viral Figure 8.8. The HIV virus. Courtesy of NIAID.

another person. The influenza virus is particularly adept at penetrating epithelial cells lining the respiratory tract, killing cells that it infects. The resulting inflammation triggers a cough reflex to clear airways of foreign invaders. During influenza infection, the immune system produces large quantities of a substance called interferon. Interferon leads to the common symptoms of the flu: fever, muscle aches, headaches and fatigue [4]. Let’s look at a more deadly pathogen – the HIV virus. HIV consists of a central core of RNA and an enzyme

enzyme reverse transcriptase uses the host machinery to copy the viral RNA into viral DNA. The viral DNA then directs the lymphocyte to produce new copies of viral protein and RNA which are assembled into protein coated viral particles within the lymphocyte. These mature particles can then bud out from the cell to release new copies of the virus into the host. Thus, HIV infection destroys an important component of the immune system; without treatment, patients develop AIDS and severe immunodeficiency. Patients with severely compromised immune function become susceptible to a variety of opportunistic infections which can result in death [3].

Figure 8.9. Life cycle of the HIV virus. NCI/Trudy Nicholson.


Biomedical Engineering for Global Health

Figure 8.11. A microscopic view of the mucous membrane of the uterine cervix.

tebrates possess an adaptive immune system, capable of recognizing and adapting itself to defend against any invader. The adaptive immune system becomes important when the innate immune system cannot defend against attack. The adaptive system also provides the immune system with “memory” [3]. Figure 8.11 shows a microscopic view of a protective physical barrier – in this case, the mucous membrane lining of the uterine cervix. This lining is about 250 microns thick (about the diameter of a human hair) and consists of multiple layers of specialized epithelial cells.

How are we protected from bacterial and viral attack? Evolution has provided two simple protective strate-

The epithelial cells at the bottom of the layer are rapidly dividing and are responsible for regenerating the epithelial tissue as it dies. As we move toward the surface of the epithelium, the cells become more mature. Cells at the very top layer are dead, but the tight junctions between these cells provide an important barrier which

gies: (1) keep pathogens out, and (2) kill them if they get in. To accomplish these goals, we are protected by three layers of immunity (Figure 8.10). First, physical barriers act to keep pathogens out. The most

is difficult for many pathogens to cross. In addition, substances present on the surface of many mucus membranes and the skin provide a chemical barrier, which functions to trap pathogens and may contain enzymes

important physical barriers are the skin and mucous membranes. These barriers must defend an enormous area – humans have over two square meters of skin and 400 square meters of mucous membranes [6]. Second, all animals possess an innate immune system to fight pathogens that make it past these physical barriers. The innate immune system recognizes molecular patterns typically associated with pathogens and responds

or other molecules with anti-bacterial activity. What happens when a pathogen is able to cross this barrier? You have no doubt experienced this if you have ever gotten a splinter (Figure 8.12). In this case, a sharp piece of wood crosses the skin, enabling bacteria to penetrate beneath the epithelial lining. What are the symptoms you experience? Frequently the area can become red, swollen, and warm. Sometimes the area will ooze pus. These symptoms are all signs that the second line of defense – the innate immune

Figure 8.10. Layers of immunity.

How the immune system fights pathogens

with a general inflammatory response to fight those pathogens which penetrate physical barriers. Third, ver-

Prevention of infectious disease


Figure 8.12. When a splinter carrying a bacterial pathogen crosses the physical barrier provided by epidermis of the skin, the innate immune response provides a second line of defense. Tissue dwelling macrophages recognize the bacteria and begin to phagocytose the pathogen. Activated macrophages secrete chemical messengers which increase blood flow, increase the permeability of blood vessels, and recruit other white blood cells to the site of the invasion. Neutrophils recruited in this manner aid in phagocytosing bacteria. This response is known as inflammation, and results in the redness, swelling, heat and pus that are sometimes present at the site of an infection. Sadava c 2008. et al. Life: The Science of Biology, Eighth Edition. Sinauer Associates, Inc. and W. H. Freeman and Co. 

system – is kicking into gear. In most cases, the innate

the splinter appear red and warm. These chemicals also

immune system can respond to the pathogen. A specialized kind of cell called a macrophage continually patrols beneath the epithelium, to detect foreign invaders and signal the immune system to respond. Macrophages are derived from monocytes, one of the five types of white blood cells. You can think of the macrophages as guards that patrol just beneath the physical barriers (skin and mucous membranes). When macrophages

cause the blood vessels in the area to become leaky. Fluid leaking out of the blood vessels produces the swelling around the area with the splinter. The chemicals released by the macrophages also recruit other white blood cells, such as neutrophils, to the site of the infection. Neutrophils aid in phagocytosing bacteria, and these cells make up the pus which can sometimes be present in the area [6].

encounter bacteria on a splinter they ingest the bacteria in a process known as phagocytosis (Figure 8.13). When macrophages are activated in this manner, they produce chemicals which increase local blood flow. It is this increase in blood flow that makes the area around

The innate immune system is primarily effective against pathogens outside of cells. How do macrophages recognize extracellular invaders as foreign? Macrophages recognize foreign invaders in two ways. There are particular molecular patterns found on the


Biomedical Engineering for Global Health

c 2004 Figure 8.13. A macrophage “eats” a bacterium. Copyright  Dennis Kunkel Microscopy, Inc.

Video microscopy showing a macrophage phagocytosing cells of the fungus Candida albicans can be seen at: When activated, macrophages recruit circulating neutrophils to the site of inflammation. The video shows neutrophils rolling along the surface of a venule; macrophages secrete chemicals which cause neutrophils to exit vessels and enter tissue. Follow the link below for video microscopy of neutrophils rolling along a blood vessel. lab goodies/lab goodies.html

surface of many pathogenic microorganisms. These signatures are recognized by receptors on the surface of the macrophage. In order to evade this route of detection by the immune system, many bacteria have evolved to hide their surface markers behind a polysaccharide capsule [3]. In some cases, proteins within the blood of the host can bind to pathogens and mark them for destruction by macrophages as a way to guide the innate immune system. In any case, if macrophages identify an invader, they become activated. Once activated they send signals to recruit other immune system cells (neutrophils), they become killers, and they activate the third line of defense, the adaptive immune system [6].

Figure 8.14. Structure of an antibody.

The adaptive immune system has two main components. The first component, called humoral immunity, relies on large proteins called antibodies (Figure 8.14) to recognize and fight pathogens outside of cells. The second component, called cell-mediated immunity, relies on several types of white blood cells to kill pathogens inside of cells. The innate immune system recognizes general molecular signatures of pathogens and provides a generalized response. In contrast, the adaptive immune system recognizes specific molecular signatures called antigens, associated with individual pathogens. The two components of the adaptive immune system accomplish the recognition using different strategies. The humoral component of the adaptive immune system relies on the chemical specificity of antibodies to recognize different pathogens. Recognition by the cell-mediated component of the adaptive immune system is facilitated by specific receptors on the surface of lymphocytes [3]. Let’s first examine how antibodies help to recognize and kill pathogens. Antibodies are Y-shaped proteins, about 12 nm in length, which are made by the immune

Prevention of infectious disease


Figure 8.15. When a B cell binds an antigen it rapidly proliferates to generate thousands of clones.

system. The bottom of the Y is known as the Fc region, while the top of the Y has two antigen binding sites

is known as clonal expansion (Figure 8.15). Effector B cells secrete antibodies which will recognize the specific

(Fab region) which can bind either to the surface of a free bacteria or virus or to the surface of a virus-infected cell [4, 7]. The free Fc region hanging off the pathogen

pathogen targeted by that B cell. Antibodies are helpful to recognize and target pathogens outside of cells. However, many viruses hide

then binds to macrophages and neutrophils and induces them to phagocytose the tagged pathogen. The Fc region can also bind to a special kind of lymphocyte known as a natural killer cell to induce destruction of the invader

inside our cells. How do we kill viruses once they are inside the cell, where antibodies cannot reach them? This process is carried out by a special class of white blood cells called T lymphocytes. T cells recognize pro-

[4]. Essentially, you can think of an antibody as a bridge between a pathogen and the tool to kill it. The Fab

tein antigens. Again, we require a way to let T cells know which cells have been invaded by viruses. Nucle-

portion of the antibody recognizes the antigen while the Fc region interacts with other components of the immune system to initiate destruction of the pathogen. Antibodies are made by a type of lymphocyte called B cells. These B cells have special receptors on their sur-

ated cells in your body have special molecules on their surface known as major histocompatibility (MHC) molecules. These molecules help your immune system recognize the cells of your body as “self” so that they do not come under attack by the immune system [7]. When

face that are designed to recognize foreign pathogens. These receptors are antibodies with their Fc region inserted in the lymphocyte cell membrane. We all have 100 million different types of B cells, each with different surface receptors. These B cell receptors are so diverse they can recognize every organic molecule; thus, they provide the ability to recognize specific pathogens [4]. Lymphocytes respond only to foreign antigens because

a virus invades a cell, fragments of viral proteins are loaded onto MHC proteins (Figure 8.16). T cells inspect the MHC proteins on the surface of your cells and use this as a signal to identify infected cells (Figure 8.16) and target them for destruction. Like B cells, when T cells bind antigen, they undergo clonal expansion [4]. The process of clonal expansion enables the adaptive immune system to have memory (Figure 8.17). The first

self-reactive lymphocytes are eliminated during their development. When a B cell binds an antigen it begins to rapidly divide and proliferate – in one week, a clone of 20,000 identical B cells can be created [6]. This process

time the adaptive immune system is activated by an antigen, your body builds up a clone of B cells and T cells. This process takes about a week [6]. After the infection is over, most of these cells die off; however, some cells,


Biomedical Engineering for Global Health

Figure 8.18. After the first exposure to an antigen, the immune system can respond much more quickly to future exposures.

Figure 8.16. An infected cell presents antigen fragments on its surface, provoking a response from T cells.

known as memory cells, remain. If the immune system is activated a second time by the same antigen, these memory cells are much easier to activate (Figure 8.18). The response of the immune system is much faster, generally so much more rapid that you don’t experience any symptoms associated with the infection [6]. This memory explains why you are “immune” to most diseases after a first exposure (think chicken pox) and is also what allows us to develop vaccines to establish this immunity in a safer way. Why then do we get the flu more than one time? Influenza virus particles are usually 80–120 nm in diam-

eter, and consist of an outer lipid envelope, an intermediate protein capsid, and a central core of RNA. There are three major types of influenza virus: A, B, C. Most serious cases of the flu in humans are caused by type A influenza. Figure 8.20 shows a schematic drawing of an influenza virus; there are two kinds of proteins found in the lipid envelope of the influenza virus. Hemaglutinnin mediates attachment of viral particles to the host cell membrane; neuraminidase mediates release of newly formed viral particles from the host cell. Type A virus contains 8 single stranded pieces of RNA [3]. The influenza virus can evade immune extinction in two ways: antigenic drift and antigenic shift. As influenza reproduces, reproduction errors occur that

Figure 8.17. After the initial exposure to an antigen, memory cells “remember” the antigen, allowing for a faster response to future exposures.

Prevention of infectious disease


Seasonal influenza and pandemics Antiviral drugs are effective against the influenza virus if taken within two days of the onset of symptoms. Many of these drugs work by hindering the change in pH necessary for influenza virus to escape the endosome following endocytosis. Other antiviral drugs block the effect of neuraminidase, to inhibit release of new viral particles. H5N1 is resistant to some antiviral drugs, but responds to others. Mortality due to influenza fluctuates seasonally. The CDC tracks mortality, to monitor for epidemics of influenza, where the mortality rate rises above the seasonally anticipated baseline level. In 2004, an influenza epidemic occurred in the USA. Pneumonia and Influenza Mortality for 122 US Cities Week Ending 5/21/2005

% of All Deaths Due to P & I


10 Epidemic Threshold


6 Seasonal Baseline 2001





4 10 20 30 40 50 10 20 30 40 50 10 20 30 40 50 10 20 30 40 50 10 20 Weeks

influenza mortality rates for years 2001–2005. Source: CDC. Figure 8.19. Pneumonia and–2005/images/bigpicurvesumary04–05.gif Three influenza pandemics have been recorded; the most deadly occurred in 1918. New strains of influenza continue to emerge, and may result in future pandemics.

change the structure of the virus. The changes are so slight that the virus is still capable of infection and reproduction but significant enough that the immune system does not possess memory for the changed virus. Mutations in the virus have resulted in a number of different strains of influenza virus, and these are characterized by differences in the two major coat proteins, such as influenza A (H2N1) [3]. This antigenic drift is why the structure of the flu vaccine changes annually. Antigenic shift refers to much larger changes in the structure of the virus. Sometimes animals can be coFigure 8.20. The influenza virus.

infected by different strains of the influenza virus.


Biomedical Engineering for Global Health IMMUNITY

PHYSICAL BARRIERS skin, mucous membranes

ADAPTIVE INNATE IMMUNITY IMMUNITY macrophages, chemical barriers (pH, lipids, enzymes) HUMORAL CELL MEDIATED B cells antibodies T cells

Figure 8.22. Overview of the components of the immune system. Figure 8.21. Significant influenza appearances since 1918. Adapted from

During viral replication, viral gene segments from different influenza strains are present inside the same cell;

tagious among birds. While it usually does not cause illness in wild birds, it can quickly devastate populations of domesticated birds – with a mortality rate approaching 100% [9]. H5N1 does not normally infect humans, but in rare cases it can be transmitted from birds to peo-

occasionally, these gene segments can randomly reassociate to create a completely new strain of the virus. Reassortment can lead to new strains of influenza that are capable of infecting humans. If a new strain can be

ple. Over the last few years, several hundred cases of H5N1 have been reported in humans, mostly in Asia. Fatality rates have been reported to be approximately 50% [12]. The majority of these cases were acquired

easily transmitted from person to person, an influenza pandemic (global outbreak) can occur because people

via bird to human transmission; however, a few cases of human to human transmission have been reported [13].

do not have any immunity to this new strain [3]. As a result of genetic shift, there have been three pandemics of flu in the past century (Figure 8.21). The largest occurred in 1918, when the Spanish flu

Today, there is considerable concern that the H5N1 will mutate into a highly infectious strain that can spread rapidly from person to person. This could lead to a pandemic with potentially very high mortality. Later we will examine the challenges associated with preventing such

(H1N1) affected between 20–40% of the world’s population, killing more than 50 million people worldwide and 675,000 in the USA [9, 10]. The virus had a particularly high mortality rate in young adults. The impact of this pandemic on the USA was so significant that life expectancy dropped by ten years [10]. Why did the Spanish flu kill so many? To gain insight into why this virus was so lethal, scientists recently reconstructed the H1N1 influenza virus and used it to infect macaque monkeys; the experiment was carried out in a biosecure facility [11]. They found that the virus triggered a severe immune reaction in the animals; the immune reaction was so severe that it provoked the body to begin killing its own cells and could rapidly destroy large amounts of lung tissue, leading to death. Today, there is particular concern regarding the strain of the flu known as avian flu, or influenza A(H5N1). This strain of flu is ordinarily carried by wild birds and does not grow well in human cells. The virus is spread from bird to bird through fecal contact, and is very con-

a pandemic.

Summary The genetic simplicity of many infectious pathogens allows them to undergo rapid evolution; our immune system must constantly cope with pathogens that develop selective advantages to avoid detection and destruction by the immune system. The complexity of the immune system is remarkable, and we have seen only an overview of how this system functions to protect us from disease (Figure 8.22). Physical barriers prevent many pathogens from entering tissue. The innate immune system provides a rapid response system to detect and attack many extracellular pathogens that make it past these physical barriers. The adaptive immune system provides the diversity to recognize over 100 million antigens and to remember which have been encountered previously. Antibody-mediated immunity can fight pathogens outside of cells, and cell-mediated

Prevention of infectious disease


Table 8.1. Advantages and disadvantages of three types of vaccines [5]. Type of vaccine



Noninfectious vaccines: Vaccinate with killed pathogen

Stimulates humoral immunity Minimal danger of infection

Does not stimulate cell-mediated immunity Usually need booster vaccines

Live attenuated bacterial or viral vaccines: Vaccinate with weakened pathogen

Stimulates both humoral and cell-mediated immunity Usually provides lifelong immunity

Poses some risk of disease, particularly in immuno-compromised host

Subunit vaccines: Vaccinate with pathogen subunit

Stimulates humoral immunity Minimal danger of infection

Does not stimulate cell-mediated immunity Usually need booster vaccines

immunity can fight pathogens within cells. In the next section, we examine how we can exploit the properties of the adaptive immune system to provide immunity to

Inactivated organism vaccine

dangerous pathogens in a process called vaccination.

has been killed or inactivated so that it can still elicit immunity, but it can not cause disease. For example, pathogens can be treated with chemicals (like formalde-

How vaccines work Vaccines are the most cost effective medical intervention known to prevent death or disease. Vaccination is the practice of manipulating the adaptive immune system to artificially induce immunity. The goal of vaccination is to stimulate the adaptive immune system to make memory cells that will protect the vaccinated person against future exposure to a pathogen, without causing the symptoms of disease. How can we stimulate the adaptive immune system to make memory cells? It is easiest to stimulate humoral immunity because this component of the immune system responds to extracellular pathogens. In order to create memory B cells, we simply need to expose B cell receptors to the pathogen or some part of the pathogen. It is more difficult to stimulate cell-mediated immunity because this component of the immune system responds to intracellular pathogens. Memory killer T cells are created only when antigen presenting cells are infected with a pathogen. Our goal in making a vaccine is to provide this exposure in a safe way. There are several types of vaccines which can stimulate the adaptive immune system to provide memory and protect against future exposure to a pathogen. Here, we will examine three types of vaccines including: (1) inactivated organism vaccines, (2) live attenuated vaccines, and (3) pathogen subunit vaccines. Table 8.1 compares the advantages and disadvantages of each of these strategies.

Generally considered to be the safest type of vaccine, noninfectious vaccines are based on a pathogen which

hyde) to kill them. When a patient is exposed to the dead pathogen, they mount an immune response without becoming infected. The Salk (inactivated) Polio vaccine, the hepatitis A vaccine, and the rabies vaccine are examples of killed pathogen vaccines [5]. The immune system encounters the pathogen outside of cells, so that this approach stimulates humoral immunity. Because the pathogen has been killed it does not infect cells, so this approach does not stimulate cell-mediated immunity. As a result, booster vaccines are usually required to maintain lifelong immunity to these diseases. When using vaccines based on an inactivated organism alone, the vaccine may not stimulate the immune system sufficiently. In order to increase the response of the immune system, the vaccine is sometimes formulated with an adjuvant – a substance that increases the response of the immune system. Aluminum salts are frequently used as an adjuvant in vaccines [14].

Live attenuated vaccines A stronger immune response can be produced by vaccinating with a live pathogen which has been weakened so that it does not cause disease, but still elicits immunity. In this approach, the pathogen is grown in host cells in the cell culture laboratory in order to produce mutations which weaken the pathogen so it cannot produce disease in healthy people, but can still produce a sufficiently strong immune response that


Biomedical Engineering for Global Health

protects against future infection. The Sabin Polio vaccine (oral Polio), measles, mumps, rubella (MMR), and varicella vaccines are examples of vaccines made using this approach [5]. This approach has a number of advantages. Live, attenuated organism vaccines contain the target antigens for humoral antibody, the pathogen molecular patterns for stimulating innate immunity, and, because of the invasiveness of the organism, it can deliver antigens effectively [14]. The immune system treats live, attenuated vaccines in just the same way as it would an infectious pathogen. Thus, these vaccines stimulate both humoral and cell-mediated immunity. As a result, they usually provide lifelong immunity. However, such vaccinations can produce disease in an immunocompromised host [5]. A technical challenge of this approach is that it is not always feasible to produce strains of a pathogen that have been attenuated sufficiently. In addition with pathogens that undergo antigenic drift, there is a finite risk of reversion to a virulent form [14]. Vaccines based on live attenuated pathogens account for approximately 10% of total vaccine sales [15].

Subunit vaccines Our immune system generally recognizes and responds to a portion of an infectious organism known as an antigen. If we can identify the antigen that will produce an immune response, we can purify that antigen and use it as the basis for a vaccine. This type of vaccine is very safe because there is no risk that it can lead to infection, even in an immunocompromised host [5]. Several strategies have been developed to produce subunit vaccines. For example, many bacteria produce disease by secreting toxins that interfere with normal cell function. We can create an immune response to these toxins by vaccinating with purified bacterial toxins that have been chemically treated to make them harmless. This type of vaccine is known as a toxoid vaccine and it produces an immune response without symptoms of disease [14]. The diphtheria and tetanus vaccines are examples of toxoid vaccines [5]. Alternatively, a subunit vaccine can use part of a pathogen to induce an immune response but not disease. Our immune system responds to the polysaccharides found on the surface of certain bacteria. We can grow bacteria in culture and extract

these polysaccharides from the culture media the bacteria are grown in to develop a vaccine. The haemophilus influenza type b (HiB), and pneumonoccocal vaccines are examples of vaccines made using this approach [5]. In a related approach, we can use the tools of genetic engineering to manufacture a pathogen protein; again exposure to the pathogen protein provides immunity without causing disease. The vaccine for Hepatitis B is based on a protein found on the surface of the virus. We can insert the genes that encode this surface protein into yeast, and use the yeast as a factory to produce the protein for the vaccine [5].

History of vaccines Throughout history vaccination has protected individuals against disease. As early as the seventh century, there are records of Indian Buddhists drinking snake venom to induce immunity (likely through a toxoid effect) [17]. In the second millennium, vaccination against smallpox was carried out in central Asia, China and Turkey. People recognized that those who had been exposed to variola (also called cowpox, which produces only mild symptoms) usually did not contract smallpox, an often fatal disease. These two antigens are sufficiently similar that exposure to one protects against the other. In 1721, the idea of variolation against smallpox moved from Turkey to England [18]. Independently in 1796, Edward Jenner, a country doctor living in England, noted the relationship between smallpox and cowpox. He observed that dairy milkmaids frequently contracted cowpox, which caused lesions similar to that of smallpox. The milkmaids who had cowpox almost never got smallpox. Jenner carried out an experiment where he injected cowpox pus into a young boy named James Phipps. He later injected Phipps with pus from smallpox sores and noted that Phipps did not contract smallpox. (We will return later to discuss the ethical problems associated with Jenner’s experiment.) Despite the ethical flaws of the experiment, it was a scientific success and Jenner was the first to introduce large scale, systematic immunization against smallpox. While his work ultimately led to the elimination of smallpox from the world, it was not immediately embraced by all. Many people were deeply suspicious of the practice of introducing animal products into

Prevention of infectious disease


Edible vaccines Subunit vaccines can also be produced in plants; genes are introduced which induce the plant to make the protein that will stimulate immunity. Usually subunit vaccines are expensive because you must purify the protein grown in culture. With plants, you don’t have to do this. Plants can be grown locally, avoiding problems with vaccine transport. Usually subunit vaccines are injected, because the digestive system will destroy the protein in the stomach before it can be presented to the immune system. However, in plants, the protein is protected by the cell walls of the plant cells. As a result, the protein survives until it reaches the intestine, where immune cells in the intestinal wall are activated. Studies have shown that tomatoes and potatoes can synthesize antigens from major causes of diarrhea: Norwalk virus, enterotoxigenic E. coli, and Vibrio cholera. Feeding these plants to animals can evoke immune response, and provide partial exposure to the real toxin. Small trials in human volunteers have produced immune reactivity in people. Scientists are grappling with the problem of getting the plants to produce a sufficient amount of antigen. When plants produce large quantities of antigen, they tend to grow poorly. Also, some plants require cooking in order to be palatable. Plants containing edible vaccines cannot be cooked because heating denatures the proteins, thereby preventing an appropriate immune response [16].

Figure courtesy of Jared Schneidman Design.

their own bodies. During the 1800s, cartoons appeared mocking Jenner and depicting the transformation of the recently vaccinated into sickly cows and fantastic beasts (Figure 8.23) [1]. Despite these concerns, rapid progress followed. In 1885, Louis Pasteur developed the concept of an attenuated vaccine and produced the first vaccine against rabies [1]. In the early 1900s, the concept of toxoid vaccines was developed, leading to vaccines for diphtheria and tetanus [5]. In the 1950s, the tools to maintain cells alive in tissue culture were developed [19]. This scientific advance led to a live attenuated vaccine for polio, a discovery for which Enders, Robbins, and Weller won the Nobel prize. Vaccines for measles, mumps, and rubella were developed in the 1960s [17].

Figure 8.23. Cartoon mocking Edward Jenner and vaccination. Wellcome Library, London.


Biomedical Engineering for Global Health

Table 8.2. The incidence of many infectious diseases in the United States was dramatically reduced by vaccines [20]. Disease

Peak # of Cases

# Cases in 2000

% Change


206,929 (1921)




894,134 (1941)




152,209 (1968)




265,269 (1952)




21,269 (1952)



57,686 (1969)




1560 (1923)



∼20,000 (1984)



26,611 (1985)



HiB Hep B

However, smallpox is also the first human disease to be eradicated from the face of the Earth by a global immunization campaign. As a result, we no longer routinely immunize against smallpox. The Jenner vaccine was first available in the early 1800s [1]. However, it was difficult to keep the vaccine viable enough to deliver in the developing world. In the 1950s a much more stable, freeze dried vaccine was developed, making it practical to deliver vaccine worldwide. In 1959, the Twelfth World Health Assembly set a goal to eradicate smallpox from the globe. However, little progress was made until


1967 when sufficient economic resources were dedicated to vaccinate at least 80% of all populations and to survey for and contain outbreaks. At that time, approximately ten million cases of smallpox occurred per year. On May 8, 1980, the Certification of Smallpox Eradica-

Table 8.2 shows the dramatic reduction in the incidence of infectious disease in the USA following rou-

tion declared the world to be smallpox free, and we no longer routinely vaccinate for smallpox [18]. In the developed world, routine childhood immuniza-

tine vaccination. For example, there were more than 200,000 cases of diphtheria in the USA in 1921, the year of peak incidence. In 2000, only two cases of diptheria were reported in the USA; the dramatic reduction in inci-

tion has dramatically reduced the incidence and mortality associated with many diseases. However, the situation is drastically different in the developing world. In 1974, only 5% of the world’s children received six

dence is due to routine childhood immunization. Polio has been eliminated in the USA due to vaccination. In

vaccines recommended by WHO [23]. At that time, the WHO set a goal to immunize at least 80% of the

2005, the CDC announced that rubella is no longer a health threat in the USA [21]. In 1965, there were 12.5 million cases of rubella in the USA. As a result, more than 12,000 babies were born deaf, blind or both and

world’s children against these six diseases by 1990 [17]. The program has been a tremendous success, and as of 2004, vaccine coverage has reached nearly 80%. As a result of vaccination, 20 million lives have been saved over the last two decades. Figure 8.24 shows the dramatic reduction in the incidence of reported cases of

6200 children were stillborn. In 2004, there were only nine rubella infections in the USA [22]. Smallpox is one of world’s deadliest diseases, having caused more deaths in history than any other disease.

measles and pertussis during this time period. While these achievements have dramatically reduced child

Measles and Pertussis Cases Reported to the WHO 5000000 4500000 4000000

3000000 2500000




1500000 1000000 500000


06 20


20 04


20 00


19 98


19 94

19 92

19 90


19 88


19 8

19 8

19 82

0 19 80

Reported Cases


Figure 8.24. The worldwide incidence of measles and pertussis have dropped drastically in response to efforts to increase childhood vaccinations. Source: WHO, Immunization surveillance, assessment and monitoring.

Prevention of infectious disease mortality, the 20% of children who do not receive these vaccines account for nearly 1.4 million preventable deaths each year due to pertussis, diphtheria, polio, measles, tetanus and tuberculosis [24].

Listen to the story of the first volunteer to test a vaccine designed to prevent Ebola in a phase I clinical trial [25]. NPR Story – Nurse Takes Plunge in Ebola Test

With new technologies, it is likely that nearly a dozen vaccines will soon be available. Recently licensed new vaccines include Gardasil (Merck) to prevent HPV infection and Prevnar (Wyeth Pharmaceuticals) for pneumococcal disease. Half of all vaccines have been developed in the past 25 years (about one per year, compared to about one every five years before this) [14].

How do we test the effectiveness of new vaccines?


trial, a larger number of volunteers (several hundred) are tested, over a period lasting a few months to a few years [28]. Generally, a phase II trial is a controlled study, with some volunteers receiving the vaccine and some receiving a placebo (or existing vaccine). The volunteers are monitored to see if they mount an immune response or contract the disease (vaccine effectiveness) or to see if they develop side effects (vaccine safety). Finally, vaccines enter phase III clinical trials involving large numbers of volunteers (several hundred to several thousand) [28]. These trials last years and are usually carried out as controlled double blind studies, with some volunteers receiving vaccine and some receiving placebo (or existing vaccine) [26]. The trial is referred to as double blind because neither patients nor physicians know which was given. If the vaccine is proven to be safe and effective in phase III clinical trials, the manufacturer can apply to the FDA to sell the vaccine. Licensure by FDA is required before a company can market the vaccine. Generally this requires about a decade [28]. Vaccines must be made

Vaccines are first tested in the laboratory to see if they initiate a response in cell culture or tissue culture sys-

following strict manufacturing guidelines and quality control procedures known as current Good Manufacturing Practices (cGMP) [26]. These regulations ensure

tems. If successful, the vaccine is then tested in animal model systems. The animal must be susceptible to infection by the agent against which the vaccine is directed

that the plant, equipment and procedures are designed to make a vaccine that is as safe as possible. Each batch of vaccine must be tested for safety, potency, purity,

and should develop the same symptoms as humans. Vaccines which are successful at preventing disease in animals can then enter human trials. We will learn more about the process of testing drugs

and a sample lot must be sent to the FDA [28]. After approval to market a vaccine is given, the FDA continues to monitor vaccine safety in a process known as post-licensure surveillance. Doctors must

and devices in patients later in Chapter 9, but briefly, these trials have three phases. In phase I trials, the vaccine is tested in a small number of volunteers (20–80 persons) [26]. Usually phase I trials are carried out in healthy adults and last a few months. The goal of phase I trials is to determine the vaccine dosages necessary to produce protective levels of immunity, as well as to evaluate side effects at these dosages. Before phase I trials

report adverse reactions after vaccination to the FDA and the Centers for Disease Control and Prevention (CDC). The reporting system is known as the Vaccine Adverse Events Reporting System (VAERS); it receives as many as 12,000 reports per year, of which 2000 are serious. Most are unrelated to the vaccine, but some can indicate rare but serious side effects that were not observed in phase III clinical trials [29].

can be carried out, the Food and Drug Administration (FDA) must approve the vaccine as an Investigational New Drug (IND) [27]. If successful results are obtained in phase I trials (immune protection with minimal side effects), then the vaccine goes into phase II clinical trials. In a phase II

How effective are vaccines? In general, about one to two of every 20 people immunized will not have an adequate immune response to a vaccine [30]. Yet, vaccination has largely eliminated many diseases. This occurs because of a phenomenon


Biomedical Engineering for Global Health Global Diphtheria Cases Reported to WHO


Diphtheria Cases Reported







known as herd immunity. Vaccinated people have antibodies against a pathogen, and as such they are much less likely to transmit that germ to other people. As a result, even people that have not been vaccinated are protected. About 85–95% of the community must be vaccinated to achieve herd immunity [1]. When herd immunity is lost, outbreaks of once uncommon diseases can occur. In the early 1990s, eastern Europe experienced an outbreak of diphtheria. Universal childhood immunization against diphtheria was introduced throughout the Soviet Union in 1958, and diphtheria incidence dropped and remained at very low levels for more than thirty years. However in the early 1990s, childhood immunization rates fell in the newly independent states of the former Soviet Union. Declining economic conditions, large population migrations and low immunization rates led to a resurgence of the incidence of diphtheria (Figure 8.25) [31]. Massive efforts to vaccinate children and adults were required to bring the epidemic back under control. As vaccines become more widely available, they have often lost their allure in many countries [1]. With the success of widespread immunization, the threat of illness that initially led to the support of vaccines has diminished. Instead, attention has become increasingly focused on the risks of vaccination. One needs only to Google the terms “vaccine safety” to see the results of this shift in public perception in the United States [32]. In 1954, more Americans knew about the field trial of the Salk Polio vaccine than knew the full name of US President Dwight David Eisenhower. At the time, the March of Dimes carried out extensive media campaigns,






























Figure 8.25. The global incidence of diphtheria, 1980–2007. Source: WHO, Immunization surveillance, assessment and monitoring.

which increased awareness of the risks of polio and efforts to develop a vaccine. As a result, most Americans understood the risks of polio and were anxious to be vaccinated [33]. As vaccination has reduced the incidence of infectious disease, groups opposing vaccines have proliferated. An increasingly cynical public sometimes regards information campaigns about the benefits of new vaccines as hype to increase the revenue of pharmaceutical companies. The Internet has facilitated the spread of dissenting views about the risks of vaccination [32]. For example, some watchdog groups have questioned the link between a rise in autism and the use of the preservative thimerosal in vaccines. A careful series of scientific studies have shown there is no link between thimerosal in vaccines and austism. Even so, the FDA ceased to license thimerosal-containing vaccines [1]. Similar claims linking autism and the MMR vaccine have arisen [1]. In 1998, the journal The Lancet published a paper that investigated the link between chronic gastro-intestinal disease and severe developmental regression and autism in a small group of children [34]. The paper noted that most instances occurred after MMR immunization, but the researchers noted they had not yet proved a causal link. Though the paper was appropriately cautious, the lead author of the study, London-based researcher Andrew Wakefield, held a press conference and warned parents that it would be safer if their children received individual vaccinations for measles, mumps and rubella, rather than the combined shot [32]. Fearing a decline in immunization rates, the UK Department of Health urged parents

Prevention of infectious disease


Table 8.3. Immunization schedule for children from birth to six years of age (courtesy of CDC).

not to reject MMR vaccinations. Between 1998 and 2004, fueled in part by inflammatory medial coverage, MMR immunization rates declined in Britain to only 80%, falling to just 62% in some areas of London. Two subsequent, larger studies in 1999 showed there was no link between the MMR vaccine and autism. In 2004, collaborators on Wakefield’s paper publicly rejected the link between autism and the MMR vaccine. That same year, Wakefield was accused of having misled the editors of The Lancet by concealing the fact that his research was partially funded by the legal team seeking compensation for parents who believed their children were injured by the MMR vaccine [32].

smallpox vaccination outweighed an individual’s right to privacy [1]. All 50 states have school immunization laws. These laws provide for exemptions based on medical reasons (50 states), religious reasons (48 states), and philosophical reasons (15 states) [28]. Tables 8.3 and 8.4 indicate the CDC recommended childhood and adolescent vaccination schedules in the USA, respectively. Children are now routinely immunized against 16 diseases (Figure 8.26) [35].

How vaccines are made There are substantial scientific and engineering chal-

Who receives vaccines that have been licensed by the FDA? Generally, recommendations are made by the Centers for Disease Control and Prevention (CDC) work-

lenges associated with developing new vaccines. For new vaccines to impact public health, we must be able to manufacture hundreds of millions of doses of vaccine. Each and every dose must be safe and effective and equivalent. Because vaccines are given to healthy

ing in conjunction with expert physician groups regarding when the vaccine should be used and who should receive it. In making recommendations, these experts weigh the risks and benefits of the vaccine, as well as the costs of vaccination [28]. In addition, some vaccinations are required by law. In 1905, the US Supreme Court ruled that the need to protect public health by requiring

children and adults, the burden of ensuring that each dose of vaccine is safe is particularly high [36]. Large scale manufacturing of vaccines involves substantial engineering challenges. It requires the ability to take a candidate vaccine developed in a basic research lab and scale up the manufacturing process to make millions of doses. As we will see, to do this effectively

Childhood illness and vaccines


Biomedical Engineering for Global Health

Table 8.4. Immunization schedule for persons from seven to 18 years of age (courtesy of CDC).

Seasonal influenza vaccine Influenza is the seventh leading cause of death in the USA. It is the leading cause of death in children aged from one to four years old, and pneumonia associated with influenza causes 90% of deaths in people over the age of 65 [15]. The Advisory Committee on Immunization Practices, a branch of the CDC, recommends an annual influenza vaccine for children between the ages of six months and five years, pregnant women, people 50 years of age and older, people of any age with cerFigure 8.26. By age two children must receive more than 20 shots. Sometimes as many as five shots are required in a single visit to the pediatrician. Courtesy of CDC/James Gathany.

requires that scientists and bioprocess engineers work closely together; typically a team of at least ten people is required to lead and coordinate such a complex project [36]. As an example of these challenges, we next consider how the seasonal influenza vaccine is made, and the hurdles that must be overcome in order to produce sufficient vaccine to prevent future influenza pandemics.

tain chronic medical conditions and people who live in nursing homes and other long term care facilities [37]. The antigenic drift of the influenza virus presents special challenges for developing an influenza vaccine. The number and type of influenza strains circulating among the population varies dramatically from year to year. Generally, it takes two to four weeks following vaccination for people to develop protective immunity [15]. Thus, to be effective the influenza vaccine must be available in advance of the peak influenza season. Because of the long lead time to produce millions of doses of the flu vaccine and the time required for vaccinated people to develop immunity, the choice of strain to be used in the vaccine must be made months in advance of flu

Prevention of infectious disease


Figure 8.27. Manufacturing process of current flu vaccine using chicken eggs. Reprinted with permission from the National Academies Press, Copyright 2000, National Academy of Sciences.

season, increasing the chances of selecting the wrong strain.

a manufacturing error at the Chiron plant, and those doses could not be sold [33].

When the influenza vaccine was first developed in the 1940s, it provided protection against only one strain of the influenza virus (monovalent vaccine). In the

As a result of these shortages, scientists have begun to examine alternative, more rapid methods of manufacturing the influenza vaccine. The most common

1960s and 1970s, vaccines were developed that protected against two strains (bivalent vaccine), increasing the chances that more people would be protected against circulating strains. Starting in 1978 and continuing

influenza vaccine used today is based on an inactivated form of the virus; chicken eggs are used as small bioreactors to grow sufficient quantities of the virus, which is then harvested and inactivated. The current manu-

to today, three strains are included (trivalent vaccine): two A strains and one B strain. Between 1970 and 2004,

facturing process relies largely on technology developed more than 60 years ago [14].

the formulation has changed 40 times for one of the strains. On eight occasions, changes have been made in two strains, and on one occasion, all three strains were changed [15]. Recent shortages in the availability of the influenza vaccine highlight the engineering challenges associated

Once a year, for each hemisphere, experts gather to decide upon the vaccine composition. This is a time consuming step, requiring about seven weeks [15, 38]. Figures 8.27 and 8.28 show the steps in the process of manufacturing the current flu vaccine. In order to produce enough vaccine, approximately 300 million

with producing vaccines. In 2003–4, two companies manufactured 83 million doses of flu vaccine for the US market: 48 million doses were made by Aventis in the USA and an additional 35 million doses were made by

chicken eggs are required; egg production occurs in parallel with strain selection [38]. After the strain of virus has been selected, a form of the actual virus to be grown in eggs must be developed through a process

Chiron in Liverpool, England. That year, the influenza epidemic started early and the media broadcast many stories describing patients who were hospitalized and died from influenza and subsequent pneumonia. The demand for vaccine exceeded supply and many people could not obtain the vaccine. The shortage of vaccine was even more dramatic the following year. Although more doses were manufactured (Aventis made 55 million doses and Chiron made 48 million doses), there was

called reassortant preparation. In this process, cells in culture are co-infected with the wild type strain and a strain which has been adapted to grow very efficiently in eggs. The goal is to create a new strain of the virus – one which is capable of producing immunity in people but will grow well in eggs. Eggs are then inoculated with the reassortant preparation in order to produce large amounts of the virus. Fluid containing the virus is then harvested from the chicken eggs, purified using


Biomedical Engineering for Global Health

Figure 8.28. Timeline of yearly flu vaccine manufacturing process. Reprinted with permission from the National Academic Press, Copyright 2000, National Academy of Sciences.

Figure 8.29. Manufacturing process used to produce virus in cultured mammalian cells as an alternate form of vaccine production [40]. From Estell, 2006. Reprinted with permission of the National Academy of Engineering.

centrifugation and filtration, and inactivated using formalin. This process occurs for each of the three strains.

be produced – all other elements in the process can stay the same from year to year. However, because the pro-

Purified, inactivated virus from each strain is then combined, and packaged into doses of the trivalent influenza vaccine [15]. Inactivated influenza vaccine is stored in the refrigerator and loses its immunogenicity if frozen [39]. After the vaccine is made, the manufacturer must still carry out phase I, phase II and phase III clinical trials to test the vaccine efficacy and safety.

cess is cumbersome and involves long lead times, there are concerns that it will not be possible to produce sufficient vaccine to prevent an influenza pandemic should a virulent new strain of flu emerge. Until recently, egg production was seasonal raising the likelihood that a pandemic might occur at a time when no eggs are ready. Because of this concern, industry has recently changed

This entire process (including licensing and safety testing) must be repeated each year. All unused vaccine is discarded. The monovalent concentrates cannot be reused after 12 months [15]. Because the yields of new strains are not known in advance, it is difficult to ensure that the manufacturing process will give sufficient quantities of a new strain. There are some important advantages of this man-

to a cycle of continuous egg production. Furthermore, because of concerns that an avian influenza virus could infect populations of chickens that produce eggs to make vaccine, flocks associated with egg production are now under strict biosafety control so they cannot be wiped out [15]. Updating the manufacturing process may increase the speed with which new vaccine could be pro-

ufacturing process. Because it has been used for many years it is well tested and understood. The only part that changes from year to year is the structure of the virus to

duced. Instead of using eggs as bioreactors to grow the influenza virus, it is possible to grow the virus in mammalian cells in culture (Figure 8.29). The process

Prevention of infectious disease


of developing the vaccine is similar, except it is made in the mammalian cells rather than in eggs. The mammalian cells are kept in a bioreactor. The cell line to be used must be able to grow the influenza virus in large numbers, and be suitable for a wide variety of flu strains. Several cell lines which meet these requirements are available and they can be grown in chemically defined, synthetic growth media. The purification and inactivation procedures are similar to those used in egg based systems. An important advantage of mammalian cell culture based systems is that it completely avoids the need to grow eggs from biosecure flocks. In addition, growing virus in this manner may lead to higher initial purity [41]. Can we switch to systems that use mammalian cell culture? In the event of pandemic influenza, cell culture based manufacturing of vaccines could provide an advantageous alternative to traditional egg-based systems. At present, there are approximately 1.5 million liters of cell culture capacity in the USA. However, this capacity is currently dedicated to the production of other essential drugs and on a very limited basis, vaccines. Given the high cost for construction and validation of a biological production facility (upwards of a billion dollars), it is not economically feasible to have production facilities sitting idle. In the event of an emergency, such as pandemic influenza, cell culture facilities would need to halt production of other essential drugs in order to accommodate the demand for vaccines [38]. Another approach to vaccine production is to use recombinant methods to produce a subunit vaccine for influenza. We have seen that the influenza virus has two major antigens on its surface: the hemagglutinin and neuraminidase proteins. The recombinant process involves taking the DNA for HA and inserting it into another type of cell which can then be used as a bioreactor to produce large amounts of the HA protein (Figure 8.30) for use in a vaccine. Yeasts are currently used to make the recombinant proteins in the hepatitis B vaccine and in the HPV vaccine [42].

Pandemic influenza vaccine Given the existing egg based production capacity, let’s examine how long it would take to produce sufficient flu vaccine for global coverage in the event of a flu

Figure 8.30. Hemagglutinin protein produced through recombinant methods for use in a subunit vaccine Isin, Doruker, Bahar. Functional Motions of Influenza Virus Heamagglutinin (HA): A Structure-Based Analytical Approach. Biophysical Journal, 2002;82(2):569–81.

pandemic. The current trivalent vaccine contains 15 µg of the HA antigen for each strain of the virus. Each year, approximately 300 million doses of the trivalent seasonal vaccine are produced. At current production capacity then, approximately 1 billion doses of monovalent vaccine could be produced. This is enough to immunize only about 1/6 of the world’s population [14]. Table 8.5 shows that it would take almost five years to produce sufficient vaccine to provide global coverage. Unfortunately, it is unlikely that we will have five years to produce vaccine in the event of an influenza


Biomedical Engineering for Global Health

Table 8.5. Current influenza vaccine production capability [14]. Reprinted with permission from Macmillan Publishers Ltd: Nature Publishing Group [14], copyright 2006. Scale-up of influenza vaccine production Production time (including lead time)

Worldwide capacity (monovalent doses of 15 µg)

Worldwide coverage (%)

1 year



2 years



4 years and 9 months



pandemic. Because antigenic shift is a random occurrence, it is difficult to predict in advance the structure of vaccines that might be protective. To get some idea of how rapidly we might need to manufacture vaccine in the event of a pandemic, we can use simple model-

Figure 8.31. Progression of infectious diseases [44].

ing techniques to predict how rapidly a pandemic might spread throughout the world once a new strain emerges. In the simplest modeling approach, we divide the population into groups, based on their disease status,

mately 1.2 days for the influenza virus [43]. The mortality rate due to influenza, w, is estimated to be 0.0005

and track how the number of people in each group changes over time (Figure 8.31) [43]. Consider a population of N people, divided into the following groups:

per day for a virulent pandemic. Thus, the change in the number of infected people versus time is given by Equation (8.2) dE = bS I − L E − wE . dt


S = the number of people susceptible to infection E = the number of infected people who are not yet contagious

The infectious period, 1/g, is approximately 4.1 days for influenza [43]. We can express the change in the number of people who are infectious versus time with

I = the number of infectious people R = the number of people who have been infected and recovered.

Equation (8.3)

When a new strain of flu emerges, the population initially is all susceptible. Over time, people acquire disease, transmit it from person to person, and either die or recover. The decrease in the number of susceptible people as a function of time is given by Equation (8.1). dS = −bS I dt


where b represents the person to person transmission rate. The time between the point of infection and the point where a person becomes infectious is known as the incubation period, 1/L. The incubation period is approxi-

dI = L E − g I − wL . dt


Finally, the change in the number of people who have recovered versus time is given by Equation (8.4) dR = gI. dt


During this period we assume that the birth rate and the death rate due to other causes are the same. We can solve this simple system of differential equations to predict the duration of an influenza pandemic. In order to solve the equations, we need to specify the initial conditions. We assume that the population is at some initial number and begins with one initial infected person; everyone else is uninfected.

Prevention of infectious disease (a)



Figure 8.32. (a) Model demonstrating spread of virus over first 200 days. (b) Model demonstrating spread of virus taking into account international patterns of population migration. Reprinted with permission from the National Academies c 2000, National Academy of Press  Sciences.

The predictions of this simple model indicate that the vast majority of cases would occur in the first

rent manufacturing capabilities, making this amount of vaccine will take more than one year [14].

200 days of an epidemic [43]. Figure 8.32 shows the predicted number of cases per day following the initial case. The predictions of this very simple model (Figure 8.32a) agree quite well with predictions of much

Economic challenges

more sophisticated models that take into account factors such as international patterns of population migration (Figure 8.32b) [43]. In addition, these predictions agree well with epidemiologic data from the 1918 Spanish influenza pandemic, and illustrate the scary prospect that in the event of a flu pandemic, vaccine will be primarily available to survivors unless improvements are made in the manufacturing process [45]. Interestingly, the models predict that it will be far more difficult to control an influenza pandemic than SARS, because a large part of the infectious period associated with influenza occurs before the onset of symptoms. A number of researchers have developed computational tools to understand the most effective approach to prevent an influenza pandemic. Efforts such as border restrictions are unlikely to be effective. Anti-virals must be given within one or two days of symptoms to have an effect. Models indicate that the best way to prevent a pandemic is to pre-vaccinate the population. These models show that you need to immunize at least 1/3 of the population if you have a vaccine of perfect efficacy in order to prevent a pandemic [43]. However, with cur-

Are there economic incentives in place to encourage the type of investment in vaccine research and development and manufacturing processes that is needed to meet global health needs? Vaccines are made by pharmaceutical companies, and pharmaceutical companies are businesses. The current economic outlook for vaccine products is not encouraging. In 1967, there were 26 companies that made vaccines used in the USA; in 2004 there were only five. Since 1998, nine of twelve vaccines recommended for children in the USA have been in short supply [33]. As a result, children were delayed in receiving vaccines that they needed, and some children never caught up. It is particularly worrisome that vaccines for seven childhood diseases have only a single manufacturer. What happens if this company experiences a business problem or production failure [1]? Today the vaccine industry faces major hurdles. The research and development process for a new vaccine is increasingly expensive, lengthy and risky. Companies must build expensive manufacturing facilities, operate within a complex regulatory environment, and deal with a growing anti-immunization movement and a surge in liability litigation. As a result of these factors, it now costs between $110–$800 million to bring a new drug


Biomedical Engineering for Global Health

to market, and typically takes more than a decade to bring a vaccine from early development to finished product launch [26]. In 2003, a report from the Institute of Medicine noted with concern the lack of financial incentives for vaccine manufacturers and called for reforms that would encourage investment [46]. In 1998, Warner Lambert stopped making Fluogen vaccine for influenza because of economic considerations and regulatory challenges [1]. Over the past 50 years, there have been many mergers in the pharmaceutics industry. Companies which previously only made vaccines now make both drugs and vaccines. In these companies vaccine products compete with potential drug products for limited research and development dollars [33]. In general, the market of a drug is much larger – vaccines are only given once, and are often purchased by the federal government. Today, 43% of childhood vaccines are purchased by the private sector. The remaining 57% are purchased through a federal contract which covers children on Medicaid or without health insurance. In 2005, the price for vaccines recommended for children before entering elementary school was $474 if purchased on the federal contract and $782 if purchased privately. Yet, every dollar spent on vaccines, saves $5.80 in direct medical costs [26]! Litigation has also played a role in increasing the costs of vaccines in the USA. In the 1970s and 1980s a series of personal injury lawsuits claiming that the pertussis vaccine resulted in complications such as sudden infant death syndrome and mental retardation were filed. Although scientific studies showed there was no link between the vaccine and these adverse events, due to litigation the cost of the pertussis vaccine increased from 17 cents per dose to $11 per dose [33]. To stem this trend, in 1986, the National Vaccine Injury Compensation Program (VICP) was established in the USA. The program is funded by a 75 cent tax on each dose of vaccine. Litigants claiming to have been injured by a vaccine must first file a claim through this program. Some injuries are automatically eligible for compensation with no need to prove that the vaccine caused the injury. If the injury is outside the rule, the claimant must prove that the vaccine was the cause of injury.

Figure 8.33. Vaccines with full potency (left) and diminished potency (right). For most vaccines, potency cannot be determined by simply looking at the vial [39].

Claimants are free to either accept or reject the decision and award of the VICP. As a condition of accepting an award from the VICP, claimants agree not to pursue further legal action against the vaccine manufacturer. The VICP has paid about $600 million for injuries caused by vaccines administered between 1988 and 2004. Only a small fraction of these funds (2%) were used to cover lawyers’ fees; the rest went to the claimant [26].

Challenges of vaccination in developing countries Developing countries now wait an average of 20 years between when a vaccine is licensed in industrialized countries and when it is available for their own populations [47]. Economic, infrastructural, and scientific hurdles all contribute to this long delay. Vaccines are complex biological substances; they can lose their potency over time. They are more likely to lose potency if exposed to temperatures which are too cold or too warm; some vaccines are also sensitive to exposure to ultraviolet light. This loss of potency is permanent and irreversible. For most vaccines, it is impossible to tell if they have lost potency simply by looking at them (Figure 8.33) [39]. If individuals are vaccinated with vaccine that has been damaged in this way, they will not have the desired immune response. The cold chain is a system that has been developed to ensure that vaccines remain potent as they make the trip from manufacturer to the patient being immunized. The cold chain has three main components – transport and

Prevention of infectious disease


Figure 8.35. Vaccine vial monitors (VVMs) are small indicators designed to irreversibly change color upon exposure to a specific heat level. Reprinted with permission from PATH and photographer Jennifer Fox. Vaccine vial monitors by TEMPTIME Corporation.

away rather than risk using inactive vaccine. A number of tools have been developed to monitor the temperature history of vaccines during the transport and storage processes. In 1996, vaccine vial monitors (VVM) were developed based on technology originally devised for

Figure 8.34. Cold chain system. This is designed to ensure vaccines remain potent during transport from the manufacturer to the patient [39].

storage equipment, trained personnel, and management procedures. As shown in Figure 8.34, the cold chain begins with the refrigerator at the manufacturing plant, extends as the vaccine is transfered to the distributor, delivered to the provider’s office and ends with vaccine administration. It is essential to maintain proper temperature at each step in the process. It has been estimated that 17–37% of providers expose vaccines to improper storage temperatures [39]. Maintaining the cold chain is a challenge in developing countries, where a lack of infrastructure can make it difficult to maintain proper storage temperatures. When workers suspect that a container of vaccine has not been properly transported or stored, they must throw it

use in the food industry (Figure 8.35). VVMs are small indicators adhered to the tops of vaccine vials; the inner square of the VVM is chemically active and changes color irreversibly with exposure to heat. VVMs can be manufactured for a variety of heat-exposure specifications. Since March 1996, all oral polio vaccine supplied through UNICEF carry VVMs, adding only pennies to the cost of a vial. As of January 2001, all vaccines supplied by UNICEF are required to have VVMs. More than 1 billion VVMs have been delivered to developing countries, and 16 of 25 UN pre-qualified vaccine suppliers include VVMs on their products. The use of VVMs has led to a significant reduction in vaccine wastage because there is an accurate record of their temperature history [48]. Freeze watch indicators (Figure 8.36) have been developed to monitor whether vaccines have been exposed to temperatures below 0 ◦ C. The freeze watch indicator consists of a small vial of red liquid contained in a plastic casing. If exposed to temperatures below 0 ◦ C for more than one hour, expansion of the liquid causes the vial to burst, releasing the red liquid [49].

214 (a)

Biomedical Engineering for Global Health (b)

Figure 8.36. Freeze watch indicators monitor whether a vaccine has been exposed to temperatures below 0 ◦ C. Courtesy of 3M.

Most vaccines must be given by injection. This is a particular challenge in developing countries, where healthcare workers may not have access to an adequate supply of sterile needles. As a result, disposable syringes are often saved and reused. It has been estimated that over 50% of injections given in developing countries follow unsafe injection practices, which can lead to the spread of blood borne diseases [50]. A simple technologic solution is now in place to address this challenge. The BD SoloShot(TM) auto-disable syringe (Figure 8.37) is designed so that when the syringe is filled to a preset level, the plunger stops and can’t be pulled back, further ensuring that the correct amount of vaccine is delivered. After one use, the plunger automatically locks so that it can’t be reused. The BD SoloShot(tm) syringe is manufactured and marketed by Becton Dickinson and Company [51]. The price of the auto-disable syringe is rapidly dropping and is within 1 cent of disposable syringes. Since its commercial introduction in 1992, more than 2.5 billion immunizations have been delivered using BD SoloShot(TM) syringes in more than 40 countries in Africa, Asia, Eastern Europe, and Latin America. UNICEF provides only auto-disable syringes to countries that request disposable syringes [50]. An alternative approach is to develop needle free methods to deliver vaccines. In developed countries, jet injector guns are used to deliver vaccines without needles. These devices rely on a liquid stream at high

Figure 8.37. BD SoloShot auto-disable syringes automatically lock c Becton Dickenson and Company. after a single use. Courtesy and 

pressure that is used to penetrate the skin [49]. The jet injector was initially developed for use in mass injection campaigns and could immunize between 600 and 1000 people per hour. From the 1950s to the 1980s they were widely used in US school immunization campaigns and throughout the developing world. However, their use was discontinued in the 1980s, when it was recognized that the multiuse jet injectors had a small risk of transmitting blood borne pathogens from one person to another. Recently, single-use jet injectors, such as Biojector2000 (Figure 8.38) have been developed. However, they are currently too expensive for developing countries [52]. The Global Alliance for Vaccines and Immunization (GAVI) is a partnership between many public and private organizations – including UNICEF,

Prevention of infectious disease


tious diseases for which no vaccine exists. The big three challenges of most importance to developing countries include vaccines to prevent HIV, malaria and tuberculosis. We now examine the obstacles which stand in the way of developing a vaccine for HIV.

Designing a new vaccine: HIV To understand the challenges involved in developing an HIV vaccine, we must first consider the pathophysiology of HIV/AIDS in more detail. An HIV infection begins when virus is deposited on a mucosal surface. An iniFigure 8.38. Jet injector guns, like this Biojector2000, deliver vaccines without needles, using a sterile, single-use syringe. Courtesy of Bioject, Inc.

tial acute infection can sometimes produce mono-like symptoms. Viral dissemination follows, and the patient will exhibit an HIV specific immune response. As the virus replicates, it destroys an important component of the immune system – a special kind of T lymphocyte called the CD4+ lymphocytes. The rate of progression of the disease is strongly correlated with viral load. Following initial infection, patients typically experience a long latent period with no clinical symptoms. Eventually, as more and more lymphocytes are destroyed, patients develop Acquired Immunodeficiency Syndrome (AIDS) (Figure 8.40). AIDS is characterized by immunologic dysregulation, accompanied by many opportunis-

Figure 8.39. By the end of 2007, US $3.5 billion had been approved for spending in countries up to 2015. Source: GAVI Alliance Secretariat, 2008.

the WHO, the Bill and Melinda Gates Foundation, members of the vaccine industry, and NGOs. GAVI was formed in 1999 to address the long delay between vaccine availability in industrialized countries and developing countries. Between 2000 and 2015, GAVI has committed more than $2.6 billion for new and underused vaccines around the world (Figure 8.39) [53]. Scientific advances that would help make more vaccines available in developing countries include the development of temperature stable vaccines, development of vaccines that required less than three doses to immunize, and the development of needle free methods to administer vaccines [52]. Despite the truly remarkable advances in public health as a result of vaccines, there are still many infec-

tic infections and cancers. The risk of opportunistic infection is correlated inversely with the number of CD4+ lymphocytes. Left untreated, the average patient with AIDS dies in one to three years [54]. HIV infection is identified by measuring whether a person is producing antibodies against the HIV virus. The virus that causes HIV and AIDS was first discovered by Robert Gallo in 1984. At that time, Margaret Heckler, then US Secretary of Health Education and Welfare, predicted that an HIV vaccine would be developed within two years. Thirteen years later, in 1997, President Clinton declared that, “an HIV vaccine will be developed in a decade’s time.” In 2003, President Bush asked congress to appropriate $15 billion to combat the spread of HIV in Africa and the Caribbean, yet there is still no vaccine available to prevent HIV [57]. See Table 8.6 for an overview of HIV vaccines undergoing clinical trials. There are many reasons that a vaccine has proven so difficult to develop. HIV represents a unique challenge; our bodies can eliminate most acute viral infections. In


Biomedical Engineering for Global Health

Table 8.6. Overview of HIV vaccines currently undergoing clinical trials (June, 2006). Data reproduced by permission of IAVI Report and VAX, International AIDS Vaccine Initiative.

Figure 8.40. The time course of the HIV/AIDS disease progression. Source: G. Pantaleo et al. Mechanisms of Disease: the Immunopathogenesis of HIV Infection. c 1993. NEJM. 328 (327–35)  Massachusetts Medical Society. All rights reserved.

contrast, our natural immune response does not destroy HIV. In fact, HIV infection results in the production of large amounts of virus, even in the presence of killer T cells and antibody. In developing a vaccine, we are faced with the challenge of trying to elicit an immune

response that does not exist in nature. Therefore, we don’t know exactly what type of immune response a vaccine should develop [58]. How does HIV outwit the immune system so successfully? As HIV replicates inside host cells, it frequently

Prevention of infectious disease


HIV testing We test for the presence of antibodies against HIV using an ELISA (enzyme linked immunosorbent assay). In this procedure, blood is taken from a person who may be infected with HIV. In the lab blood is added to laboratory HIV virus. HIV antibodies from the blood, if present, attach to HIV antigens. Next a chemical that attaches only to antibody/antigen complexes is added. If the solution changes color, the person is making antibodies against HIV. The advantage of ELISA tests is that they are very sensitive, meaning they don’t often miss disease if it is present; the disadvantage of ELISA tests is that they are not very specific, meaning that they sometimes generate a falsely positive result in someone who does not have disease. Thus, a positive ELISA tests requires another test to confirm the presence of disease. The second test is called a western blot. Western blots are not as sensitive, but are more specific. Together, these two tests reduce the rate of false positives to 1/250,000. If this combination of HIV tests is positive, it indicates that the person is infected with the HIV virus, though he or she may not have AIDS yet. A negative ELISA test indicates either that a person is not infected with HIV, or that the person is infected with HIV but not yet making detectable level of antibodies. In general, it may take one to three months (and in rare cases six to twelve months) before the body makes enough antibodies to be detected by an ELISA [55, 56].

Figure 8.41. Ways in which HIV can undergo mutation. From [59]. Reprinted with permission from Macmillan Publishers Ltd: Nature Reviews Immunology, copyright 2006.

undergoes mutation. Many researchers believe that it is this continuous mutation of the HIV virus that enables it to escape destruction by the immune system. A high mutation rate increases the probability that a new form of the virus will emerge with a genetic advantage that enables it to survive. Figure 8.41 illustrates the ways in which the HIV virus can undergo mutation. As HIV replicates, it uses an enzyme called reverse transcriptase to copy its RNA into double stranded DNA. This DNA is inserted into the host chromosome, where it then directs production of more viral proteins that ultimately assemble into new viral particles. The HIV reverse transcriptase does not proofread this reproduction process, and on average each time the enzyme copies RNA into

Courtesy of the Nevada Department of Agriculture.

DNA, the new DNA differs at one base site [59]. HIV is the most variable virus known [60]. Additionally, if


Biomedical Engineering for Global Health an HIV vaccine must generate both cell-mediated immunity and antibody-mediated immunity [58]. What are the design goals for an HIV vaccine? A successful vaccine must produce both antibody-mediated immunity and cell-mediated immunity against multiple forms of the virus. To develop antibody-mediated immunity, the immune system must see virus or viral

Figure 8.42. Individual strains or clades of HIV. From [59]. Reprinted with permission from Macmillan Publishers Ltd. Nature Reviews Immunology, copyright 2006.

two genetically distinct forms of the HIV virus with different genetic sequences infect the same cell, DNA from both can integrate into the host genome and produce viral RNA. When new viral particles are packaged, RNA from the different parent viruses can combine to give rise to forms of HIV with entirely new genomes. HIV replicates at a very high rate, so the odds are high that useful mutations will occur over time. Over a ten year period, thousands of generations of viral reproduction have occurred; during that ten year period, the virus can undergo as much genetic change as humans would undergo in millions of years [60]! As a result of this high rate of mutation, there are many forms of HIV which a vaccine must provide protection against. The two major types of HIV are HIV1 and HIV-2. HIV-1 causes a more serious form of the disease and is responsible for the majority of HIV disease throughout the world. There are three main groups of HIV-1; more than 90% of HIV-1 disease is caused by group M HIV-1 [61]. Within this group, there are genetically distinct individual strains, called clades (Figure 8.42). It is thought that each strain may require a different vaccine [62]. Finally, development of a vaccine is complicated because there are many routes of transmission for HIV, including sexual contact and contact with contaminated blood. HIV can be transmitted by contact with virus alone or by contact with cells infected with the virus. Recall that cell-free virus is recognized and eliminated by antibodies, while cells infected with virus are recognized and eliminated by cell-mediated immunity. Thus

debris. To produce cell-mediated immunity, HIV viral proteins must be presented to immune system on MHC receptors. We have seen three strategies for developing vaccines thus far: inactivated organism vaccines, subunit vaccines and vaccines based on live, attenuated pathogens.

Noninfectious HIV vaccine strategies A noninfectious vaccine, made using killed virus or a viral subunit, will only stimulate antibody-mediated immunity, and thus will not meet the design goal. Animal trials with inactivated whole virus have shown only antibody-mediated immunity to a small number of HIV viral subtypes [63]. Similarly, trials of viral subunit HIV vaccines have shown modest antibody-mediated immunity effective against a limited number of HIV strains. One type of subunit vaccine has advanced to phase III clinical trials; this vaccine is based on the gp120 protein found in the envelope of the HIV virus. The gp120 protein is needed for HIV to enter cells. Researchers theorized that if patients could make antibody against this protein it could prevent free virus from infecting the patient’s cells. In animal models and phase I clinical trials, it has been shown that this vaccine does elicit production of antibodies against gp120. The antibodies produced neutralized HIV in a test tube. But they only recognized strains of HIV similar to those used to generate the vaccine [58]. As we have seen, the HIV virus is notoriously susceptible to mutation, and these mutations change the structure of the gp120 protein over time so that the antibody may no longer be effective against it. Despite poor results in animal trials, gp120 subunit vaccines have progressed to human clinical trials. The company Vaxgen developed a subunit vaccine based on the gp120 protein from two clades of HIV-1; the vaccine (called AIDSVAX) entered phase III clinical trials

Prevention of infectious disease


Rural village outreach: June 15, 2007 Christina Lesotho We went to a rural village where our mentor, Dr. Dudley, had promised the family of a little girl that had passed away that she would return and give back their records, etc. We got there and were greeted by the grandmother who was taking care of the little girl and seemed to be the one in charge around that small corner of the village. She had gathered other children to be tested for HIV, and soon, her little single room Basotho (what they call people or things that are from Lesotho) hut became a testing center for a few different families that came by. A social worker with us performed his first pre-testing counseling, HIV tests, and post-testing counseling for each person tested, and it was interesting to see the reactions to his explanations and what the people being tested were and weren’t comfortable discussing related to HIV and its transmission. I could not believe this was my first time seeing an HIV test kit. Sophie and I stepped outside for a moment to get some fresh air and see all the children that had gathered. They were all friendly and as soon as the camera came out, the requests for their picture to be taken did not end. They were especially interested in seeing the shots on the screen after they were taken. A few of the older girls spoke English well and one of them was telling me about her interest in school, science, social studies, and traveling. She said she wants to travel to so many countries in Africa and beyond and she hopes to become a nurse and treat others around the world. I encouraged her to return to Lesotho, of course, and kept asking her about her future schooling. She then surprised me with the fact that she cannot pay for high school, so next year may be her last year of school. I could not believe the fees for high school were so high and I really got aggravated by this. I am almost certain it is to keep too many kids from qualifying for university since the government provides college scholarships for most students who make it to that point. The ride to and from the village was really beautiful and it has been so nice not being in a busy city or hectic area. Back at the clinic, we were having lunch and some of the women on the staff at the reception and in social work/counseling started talking about handouts and food assistance and were adamantly against it. They were explaining to us that a mother came in earlier today and started crying when her child tested negative for HIV because that meant she would not get the food supplements the clinic gives to patients on medication. I was so amazed and terrified by this thought. The staff member went on to explain the extreme economic and social problems created by WFP, they call it (World Food Program, I think) and how it has decreased productivity since people have this food source to turn to. She talked about the amount of excess WFP food in villages that she has seen at funerals and other village events, and how terrible it is that the country is being destroyed by foreign aid. I was completely frustrated at the thought. I feel settled and like I am learning so much each day. I enjoy the people around me and have had an interesting time learning Sesotho words/phrases today. Hopefully I can try out some of my introductions tomorrow.

in 1998; results were announced in 2003. More than 5000 volunteers participated in the randomized, double blind, placebo controlled trial. At the beginning of the trial all were HIV negative; 3330 volunteers received

the placebo. 5.7% of those receiving the vaccine developed HIV, whereas 5.8% receiving placebo developed HIV. The difference was not statistically different. Most researchers believe that the effectiveness of this vaccine

AIDSVAX, and the remainder received the placebo. After three years, researchers compared the number of HIV infections for those receiving the vaccine and

in preventing HIV infection was limited because it does not induce cellular immunity and provides antibodies against a limited number of HIV strains.


Biomedical Engineering for Global Health

Researchers and public health workers have expressed concern that a vaccine with limited efficacy could actually increase the rate of new HIV infections. They fear that people who receive the vaccine will engage in riskier behaviors if they believe they are protected by a vaccine [62]. Such behavioral changes could possibly negate the benefit of the vaccine with limited efficacy. It is currently not known whether vaccines that do not prevent infection will delay disease progression in infected individuals. This could be an important benefit, especially in parts of the world where access to HAART is limited [59]. Furthermore, it is not known whether such vaccines could reduce viral loads in infected individuals, and this could help curb the spread of disease [64–66].

Live attenuated HIV vaccine strategies

DNA vaccines An interesting new approach to develop vaccines that stimulate both antibody- and cell-mediated immunity is to directly inject DNA that codes for viral protein into a patient. This leads the host cells to produce the protein that the DNA codes for; it is processed and loaded onto the MHC receptors and stimulates cell-mediated immunity, without any danger of causing infection. DNA vaccination approaches have shown very successful results in animal trials, generating a strong cell-mediated immune response [57]. Somewhat less successful results have been reported in human trials, where much larger quantities of DNA must be injected to generate immune response. While, many DNA based vaccines are currently in clinical trials, there are also concerns that scientists will not be able to identify a single protein that will elicit immune response against many HIV strains [62].

As we saw earlier, an advantage of vaccines that are based on live attenuated forms of a pathogen is that they stimulate both antibody- and cell-mediated immunity. Vaccines made using this approach stimulate both B cells and killer T cells and this approach is the

Carrier vaccines

most likely to stimulate the necessary immune response. However, because the HIV virus mutates so rapidly, this approach presents unique potential dangers. Because the HIV virus mutates constantly, there is a chance that

terium that does not cause disease to carry the viral genes of interest to the host cell. Again, protein is produced by the host cells and loaded onto MHC receptors where it stimulates cell-mediated immunity. This

the attenuated form of the virus used in a vaccine could undergo a mutation that restores its strength. If this

approach stimulates both humoral and cell-mediated immunity without the danger of real infection. How-

occurs, the consequences would obviously be devastating for the person receiving the vaccine. Vaccines based on a live attenuated form of the simian immunodeficiency virus (SIV) have been tested in macaque monkeys. SIV infects monkeys and is closely related to HIV. The vaccine successfully protected the animals

ever, immunocompromised individuals can become ill from the carrier. A limitation of the approach is that the carrier must be one that individuals are not already immune to [62]. For the same reasons, booster vaccines cannot be made with the same carrier. Promising results have been observed using what is

when they were exposed to SIV. However, many of the animals vaccinated progressed to AIDS like symptoms, even when not exposed to the wild type virus, although more slowly than those infected with unaltered virus

known as a prime/boost strategy (Figure 8.43). In this approach, a prime vaccine is first given using a carrier to stimulate cell-mediated immunity. This is followed by a boost vaccine using a subunit vaccine to further

[58]. Thus, new vaccine strategies are required to develop vaccines to prevent HIV infection. In the remainder of this chapter, we consider several new approaches under investigation.

stimulate antibody-mediated immunity. Phase III clinical trials of such a strategy began in 2003 using a canarypox vector to deliver HIV-1 genes that code for several proteins, followed by a boost with the AIDSVAX gp120 subunit vaccine [59].

The strength of the immune response elicited by a DNA vaccine can be strengthened by using a virus or bac-

Prevention of infectious disease


Figure 8.43. Prime/boost vaccine strategy designed to initiate both antibody- and cell-mediated immune responses against c 2002 Terese Winslow. HIV [62]. 

contributions of thousands of volunteers who are willing to participate in clinical trials. Scientific and medical progress rely on the sacrifices that healthy people are willing to make for the sake of research – and these risks are real. For example, one risk of participating in an HIV vaccine trial is that it may cause future HIV tests to be positive. This is because rapid tests mea-

Clearly, developing a new HIV vaccine will rely on the

get health insurance, obtain employment, or to travel to some foreign countries. Society has an obligation to provide protection for these volunteers. For example, we can pass laws to require insurance companies to change their screening procedures. If an applicant tests positive on an HIV rapid test, they can be required to administer a more sensitive and specific Western blot to discern between false positive rapid test and true HIV

sure antibodies against HIV and the vaccine may cause people to produce antibodies. Thus, a person participating in a vaccine trial may test positive for HIV, even though they do not have HIV. False positive HIV

infections [47]. While many volunteers willingly participate in clinical trials, public mistrust has made it more difficult to carry out large trials of candidate HIV vaccines. There is

test results can lead to social stigma. People with positive HIV tests are not allowed to donate blood. Such results can also interfere with the ability of people to

limited public knowledge about HIV vaccine research, and many people harbor suspicion that HIV vaccine research is not being carried out for the greater good.


Biomedical Engineering for Global Health

Rabid dogs and AIDS parades: June 6, 2007 Tessa Swaziland I’ve been meaning to blog about the impressive awareness about HIV/AIDS. I have no idea the extent of HIV/AIDS education (i.e., how much Swazis know about how HIV/AIDS is spread and why adherence to treatment is important) but everyone knows it’s a problem. Last weekend, when Dave and I were picking up some fruit from the market (where grandmothers, also called “gogos,” sell produce), we were briefly impeded by a parade of young adolescents holding up signs with various slogans: “No balloon; No party!” and other similar phrases. Two of my favorite campaigns are “I love you, positive or negative” and “I’m over it.” There is a huge stigma against people with HIV here, and the “positive or negative” campaign is geared toward that problem. Even the condoms say, “I love you, positive or negative.” (I only know that because there are condoms in random places around the clinic.) The “I’m over it campaign,” is pretty entertaining. There are billboards in Mbabane that have virtual text message conversations that go something like this: “My wife’s at work. Wanna cum work on me?” “No. Thanks. I’m over it.” There are several different ones. That’s the one that really stuck in my mind though. They’re pretty funny. I think the awareness is restricted mostly to the cities though. My guess is that in rural areas, people probably know a lot less. That reminds me . . . education here is not free. Many people can’t afford to send their children to school. I was actually talking to my friend Treasure (she works in administration here at the Baylor clinic) about her education experience. She was very lucky and had been sponsored by a woman in California since she was three. The mystery woman paid for all the school fees, uniforms, books, etc. Then Treasure did well enough on her exams to get a scholarship for a university.

The National Institutes of Health conducted a survey of public knowledge and attitudes regarding HIV vaccine research in 2001. Nearly half (46%) of the population surveyed strongly agreed, somewhat agreed or did not know when asked if there was already a vaccine for HIV that was being kept a secret. An even higher proportion of minorities (72% of African American and 49% of Latinos) answered this way. When asked if HIV vaccines being tested could give a person HIV, 69% of the general population strongly agreed, somewhat agreed or did not know. In Chapter 9, we consider the ethical dilemmas that arise in research involving human subjects and we outline the framework of requirements that we have put in place to ensure that such research provides an appropriate balance between the risks and benefits of research.

Bioengineering and Global Health Project Project task 4: Define the problem that your design will address In this task, you will need to be much more specific about the particular problem you are trying to solve. For example, you may have identified the need for better treatments for tuberculosis in Project tasks 1–3. In this task, you may want to consider the particular problem of developing a treatment for tuberculosis that increases patient compliance. Turn in a one-page summary of the specific health need that your design will address.

Prevention of infectious disease

Homework 1. The immune system. a. What is an antibody? Describe its structure and its function in the immune system. b. Explain the term “immunologic memory.” c. Describe the cellular-level processes that enable the adaptive immune system to have immunologic memory. 2. When you get a splinter in your toe, the area can become red, hot, swollen and ooze pus. Describe the specific causes of each of these symptoms. 3. Name the three general types of immunity and give an example of each. 4. Most common anti-HIV drugs work by inhibiting key steps in viral uptake and reproduction. a. Make a drawing which shows the major steps that occur when HIV infects a CD4+ lymphocyte. Indicate on your figure where in the viral life cycle the following classes of drugs act: (1) fusion inhibitors, (2) reverse transcriptase inhibitors and (3) protease inhibitors. b. Beginning in the mid 1990s, an increasing number of HIV-infected individuals began a drug regime called highly active anti-retroviral therapy (HAART), a combination of three or more anti-HIV drugs taken at the same time. Why is taking a combination of drugs, each targeted against a different aspect of the viral life cycle, so much more effective than taking a single drug? 5. When a TB skin test is performed, a small amount of harmless TB antigen is injected under the skin. The patient monitors for redness and swelling at the site of injection. If a patient has been previously exposed to TB, but does not currently have an active TB infection will redness and swelling be observed? Why or why not? 6. Oh no! You return to Student Health two days after receiving a routine PPD skin test. You have a red bump on your forearm that measures 12 mm in


diameter. Every year up until now, your test had been negative. a. How does the PPD skin test work, and why does a red bump form for individuals infected with TB? b. Assuming you have no significant health problems, what are the odds that the bacterium will remain in a latent, inactive state for the rest of your life? c. The PPD skin test is imperfect. Describe one instance in which the PPD skin test fails by giving a false-negative result, and describe another instance in which the test fails by giving a false-positive result. Why does the test fail in each circumstance? 7. If you are exposed to the varicella virus as a child and have not been vaccinated, you will likely develop chicken pox. If you are exposed again as an adult, you probably will not develop the disease again. a. At first exposure, what type of immunity fights off the varicella virus? b. The varicella vaccine contains a live virus. Is this safe? Why or why not? What is the advantage of this type of vaccine over a vaccine made of a dead virus? 8. A 24-year-old HIV-positive man is hospitalized because he developed pneumonia. The doctor starts the patient on antibiotics and measures the number of CD4 helper T cells in the patient’s blood. The patient has a low CD4 count. a. What are two of the three major transmission routes by which this man might have become HIV-positive? b. The doctor then performs a test and finds that the man’s serum is positive for antibodies to gp41 and gp120, the HIV envelope glycoproteins. Name and briefly describe this test the doctor ordered. 9. Answer the following questions about pathogens, the immune system, and vaccines. a. Check to indicate pathogen type(s) for which each statement applies.


Biomedical Engineering for Global Health Trait



Uses host cellular machinery to reproduce Can be killed or inhibited by antibiotics Short pathogen peptide sequences are displayed in MHC surface receptors. Living cells, usually having both a membrane and cell wall Protein capsid houses nucleic acid core Can reproduce without a host Tens of nanometers in size

b. How can T cells identify cells infected with viruses? c. Antigen binding to B-cell surface receptors and interaction with activated helper T cells activates B-cells to produce and secrete antibodies. Compare the onset and magnitude of the B-cell response for primary (initial) and secondary (subsequent) exposure to a particular antigen. d. Identify the following vaccine types from the descriptions provided. Which one is likely to confer lifelong immunity? : The pathogen is treated with chemicals or irradiated. The early version of the polio vaccine and the rabies vaccine are examples. : Mutations have been introduced to the pathogen. This form is used to prevent measles, mumps and rubella. 10. The incidence of many diseases has been reduced by widespread vaccination. However, vaccines are not available for some diseases. a. Name three diseases for which vaccines are most critically needed to improve world health. b. For one of the diseases you listed in part a, explain the major scientific and economic challenges associated with developing a vaccine.

11. Technologies for vaccine development and delivery are considered among the top ten biotechnologies that may improve health in developing countries. Imagine that you are a member of GAVI evaluating new vaccination strategies for adoption by the organization. You are asked to choose between an oral live attenuated (Sabin) and an injectable inactivated (Salk) polio vaccine for use in sub-Saharan Africa. Polio is a viral disease that can produce paralysis. It is passed through fecal–oral transmission. Assume that the two vaccines have equal efficacy in preventing polio infection. a. What does GAVI stand for? b. Place the following stages of viral infection and replication in the correct order. — Synthesis of viral proteins — Viral budding or cell lysis — Endocytosis/injection of viral contents — Binding to cell membrane c. How does a live attenuated vaccine differ from an inactivated vaccine? d. Which of the following components of the immune system must a vaccine stimulate? i. macrophages ii. B cells and T cells iii. neutrophils iv. innate immunity v. complement 12. It has been shown that unvaccinated contacts of babies who receive the Sabin Polio vaccine will develop antibodies to the virus, while unvaccinated contacts of babies who receive the Salk Polio vaccine will not. a. Explain why this might be the case. b. List two reasons why the Sabin vaccine might be preferable to the Salk for use in sub-Saharan Africa. c. What is the main risk of using the Sabin vaccine in an immunocompromised population? d. There has been much interest in eliminating poliovirus worldwide. What is the only infectious disease that has been eradicated to date?

Prevention of infectious disease


e. Discuss why the use of a vaccine led to the eradication of this disease while other diseases for which vaccines exist have not been

state’s most reported infections of the highly contagious disease, also known as pertussis, so far this year.

eradicated. f. List two properties that are necessary in order for a disease to be eradicable. g. Which vaccine would you recommend that the GAVI adopt? Name two reasons why. h. Why has there been so much focus on and investment in vaccination as a strategy in world health?

The Austin/Travis County Health and Human Services Department reported 58 confirmed cases of pertussis since Jan. 1, an unusually high number. The county has not had a whooping cough death since 2003. That year, infant Serena King died of the illness, which causes a violent cough followed by a whooping sound.

13. Portions of the following article appeared in the Austin American Statesman on May 10, 2005. Please read the article and answer the following questions. Questions about pertussis article: a. The article states that the pertussis vaccine does not protect 15–20% of children who receive it. Discuss how the concept of “herd immunity” will protect these children. What fraction of the population must be vaccinated to achieve ‘herd

King was younger than 2 months, the age at which babies get their first pertussis vaccination, when she died. State health officials are awaiting confirmation of a suspected pertussis death this year, but the patient was not from Central Texas, said Rita Espinoza, an epidemiologist at the Department of State Health Services. Pertussis is on the upswing nationally, and if current trends continue in Texas, 2005 could be

immunity’? b. The article describes a new booster vaccine called Boostrix. It states that the new vaccine

one of the worst years since vaccines have been available. As of April 30, the state had a preliminary

may be available commercially next month. What process will the FDA use to ensure that the vaccine is safe after it is approved for general use? Why is this process necessary?

count of 269 pertussis cases, compared with 192 during the same period a year earlier, according to the state health department. The worst year for whooping cough since the introduction of

c. Pertussis is generally a mild disease in adults and older children. What arguments would you make in support of widespread distribution of the booster vaccine? Travis County investigating outbreak of whooping cough County leads state in number of cases; State

vaccines in the 1940s was 2002, when the state reported 1,240 pertussis cases, Espinoza said. Health officials are worried. In 2004, the preliminary count was 1,174 whooping cough cases statewide, compared with 670 in 2003. Travis County reported 97 cases in 2004 (the state’s count for Travis County was higher, at 125; the two will

also could have another bad year for pertussis By Mary Ann Roser AMERICAN-STATESMAN STAFF Tuesday, May 10, 2005 Local health officials are investigating an outbreak of whooping cough as Travis County

reconcile the numbers later), and the county had 62 whooping cough cases in 2003, said Dr. Adolfo Valadez, the health authority for the Austin/Travis County department. “It’s a concern all over the state,” Espinoza said. “I was just down in the Valley last week,

copes with the bleak distinction of having the

and I was informed of 20 to 25 cases in an area


Biomedical Engineering for Global Health where we usually don’t hear of that many. We need to find a way to curb the cycle.” The outbreak is a warning to parents to keep their children’s immunizations up-to-date, Valadez said. It has picked up steam in the past five to six weeks, and most of the cases are in babies younger than a year old and children from ages 10 to 15, Valadez said. Schools in Austin and Pflugerville are seeing sporadic cases, but “no schools have had to be closed,” he said. “Quite, honestly, we’re looking forward to school ending. That’s how it spreads.” Espinoza and Valadez said other factors could be contributing to the uptick in cases in recent years: growing awareness of pertussis, a quicker test to diagnose the illness and waning immunity from the whooping cough vaccine. The vaccine has been changed to reduce some side effects, which could cause immunity to wear off in less than five to 10 years, Espinoza said. Also, the vaccine is far from foolproof. It does not protect 15 percent to 20 percent of the children who get it, which means adolescents can get pertussis and spread it to young children and babies who are at greatest risk of serious illness. A week ago, the Food and Drug Administration approved the use of a pertussis booster vaccine, Boostrix, for children from ages of 10 and 18. Valadez said it is expected to be available commercially as early as next month, and he was encouraged that the tool was coming to the public health arsenal. Now, children are vaccinated for pertussis at 2 months, 4 months, 6 months and between 15 months and 18 months, with a booster between ages 4 and 6, Espinoza said. Pertussis bacteria live in the nose, mouth and throat and escape into the air when people sneeze, cough and talk. The disease is usually mild in older children and adults but can cause breathing problems, pneumonia and swelling of the brain. It begins like a cold, with a mild fever

and cough, which slowly worsens and leads to coughing fits that sometimes end in vomiting. 14. Google the terms: a. Vaccine and safety b. Vaccine and dangers Do you think the sites that pop up on the two searches contain accurate health information? Why or why not? If you were a pediatrician, what would you tell the parents of your patients who had performed similar searches? A short paragraph is sufficient. 15. You have been asked to write a 500–525 word column on the Avian influenza situation for the BIOE Tribune. Your editor informs you that you must write a critique of the US plan in case of an Avian influenza pandemic. Your critique should include the scientific, economic, and public health aspects of this plan. Other topics, including potential vaccine strategies, may be addressed as well. The CDC website avian/ may provide information which will be helpful in completing this assignment. REMEMBER this is for a newspaper so make it compelling and make it interesting, but also make it TRUE! 16. Who sings “The Avian Flu . . . a three minute summary”? (Hint: she also sings “King of the Rollerama” and “The Great Metric Threat of 79.”)

References [1] Stern AM, Markel H. The history of vaccines and immunization: familiar patterns, new challenges. Health Affairs (Project Hope). 2005 May–Jun; 24(3): 611–21. [2] WHO. Mortality: Revised Global Burden of Disease (2002) Estimates. Geneva: World Health Organization; 2002. [3] Kasper DL, Braunwald E, Fauci AS, Hauser SL, Longo DL, Jameson JL, eds. Harrison’s Principles of Internal Medicine. 16th edn. New York: McGraw-Hill; 2005. [4] Sompayrac L. How Pathogenic Viruses Work. Sudbury, MA: Jones and Bartlett Publishers 2002. [5] Centers for Disease Control and Prevention. Epidemiology and Prevention of Vaccine-Preventable Diseases. 10th edn. Washington DC: Public Health Foundation; 2007.

Prevention of infectious disease [6] Sompayrac L. How the Immune System Works. 2nd edn. Boulder Blackwell Publishing; 2003. [7] Silverthorn DU. Human Physiology : an Integrated Approach. 3rd edn. San Francisco: Pearson/Benjamin Cummings; 2004. [8] Centers for Disease Control and Prevention. 2004–05 U.S. Influenza Season Summary. Atlanta, GA; 2005 July 5. [9] WHO Regional Office for the Eastern Mediterranean. Avian Influenza (Bird Flu): an Introduction. Division of Communicable Disease Control Newsletter. 2005 November (7). [10] Tumpey TM, Basler CF, Aguilar PV, Zeng H, Solorzano A, Swayne DE, et al. Characterization of the reconstructed 1918 Spanish influenza pandemic virus. Science (New York, NY). 2005 Oct 7; 310(5745): 77–80. [11] Kobasa D, Jones SM, Shinya K, Kash JC, Copps J, Ebihara H, et al. Aberrant innate immune response in lethal infection of macaques with the 1918 influenza virus. Nature. 2007 Jan 18; 445(7125): 319–23. [12] Cumulative Number of Confirmed Human Cases of Avian Influenza A/(H5N1) Reported to WHO. 2007 April 11 [cited 2007 April 23]; Available from: country/cases_table_2007_04_11/en/print.html [13] World Health Organization. Avian Influenza Frequently Asked Questions. 2005 December 5 [cited 2007 April 23]; Available from: avian_influenza/avian_faqs/en/index.html [14] Ulmer JB, Valley U, Rappuoli R. Vaccine manufacturing: challenges and solutions. Nature Biotechnology. 2006 Nov; 24(11): 1377–83. [15] Matthews JT. Egg-based production of influenza vaccine: 30 years of commerical experience. The Bridge. 2006; 36(3): 17–24. [16] Langridge WHR. Edible vaccines. Scientific American. 2000; 283(3): 66–71. [17] Bloom BR, Lambert PH, eds. The Vaccine Book. San Diego: Academic Press; 2003. [18] Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. Smallpox and its Eradication. Geneva: World Health Organization; 1988. [19] Smith KA. Medical immunology: a new journal for a new subspecialty. Medical Immunology. 2002 Sep 30; 1(1): 1. [20] Abbas AK, Lichtman AH. Basic Immunology: Functions and Disorders of the Immune System. 2nd edn. Philadelphia: W.B. Saunders; 2004. [21] Mahmoud A. The vaccine enterprise: time to act. Health Affairs. 2005; 24(3): 596–7.


[22] The infrastructure for vaccine development. Health Affairs. 2005; 24(3): 598. [23] Levine R. Millions Saved: Proven Successes in Global Health. Washington D.C.: Center for Global Development; 2004. [24] Unicef. Immunization Summary 2006. New York: Unicef and WHO; 2006 January. [25] Nurse Takes Plunge in Ebola Test. All Things Considered: National Public Radio; 2003. [26] Orenstein WA, Douglas RG, Rodewald LE, Hinman AR. Immunizations in the United States: success, structure, and stress. Health Affairs (Project Hope). 2005 May–Jun; 24(3): 599–610. [27] Vaccine Education Center – Children’s Hospital of Philadelphia. Frequently Asked Questions. 2003 [cited 2007 April 25]; Available from: consumer/jsp/division/generic.jsp?id=75743 [28] National Network for Immunization Information. Why Immunize? How Childhood Vaccines are Selected for Routine Use. 2004 [cited 2007 April 25]; Available from: howVaccines_selected.cfm [29] National Network for Immunization Information. Why Immunize? Monitoring Vaccine Safety. 2007 [cited 2007 April 25]; Available from: monitoringSafety.cfm [30] National Network for Immunization Information. Why Immunize? How Vaccines Work. 2007 [cited 2007 April 25]; Available from: parents/howVaccines_work.cfm [31] Centers for Disease Control and Prevention. Diptheria epidemic – new independent states of the former Soviet Union, 1990–1994. Morbidity and Mortality Weekly Report. 1995; 44(10). [32] Colgrove J, Bayer R. Could it happen here? Vaccine risk controversies and the specter of derailment. Health Affairs (Project Hope). 2005 May–Jun; 24(3): 729–39. [33] Offit PA. Why are pharmaceutical companies gradually abandoning vaccines? Health Affairs (Project Hope). 2005 May–Jun; 24(3): 622–30. [34] Wakefield AJ, Murch SH, Anthony A, Linnell J, Casson DM, Malik M, et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. The Lancet. 1998 Feb 28; 351(9103): 637–41. [35] Centers for Disease Control and Prevention. Childhood & Adolescent Immunization Schedules 2007 [cited 2007 April 27]; Available from: recs/child-schedule.htm#presentation


Biomedical Engineering for Global Health

[36] Buckland BC. The process development challenge for a new vaccine. Nature Medicine. 2005 Apr; 11(4 Suppl.): S16–19. [37] Smith NM, Bresee JS, Shay DK, Uyeki TM, Cox NJ, Strikas RA. Prevention and control of influenza recommendations of the Advisory Committee on Immunization Practices (ACIP). Morbidity and Mortality Weekly Report. 2006 July 28; 55(RR-10). [38] Scannon PJ. Pharmaceutical preparedness for an epidemic. The Bridge. 2006; 36(3): 10–16. [39] Centers for Disease Control and Prevention. Vaccine Storage and Handling Toolkit. 2005 June [cited 2007 April 30]; Available from: [40] Estell D. Adapting industry practices for the rapid, large scale manufacture of pharmaceutical proteins. The Bridge. 2006; 36(3): 39–44. [41] Rappuoli R. Cell culture based vaccine production: technological options. The Bridge. 2006; 36(3): 25–30. [42] Shaw A. Alternative methods of making influenza vaccines. The Bridge. 2006; 36(3): 31–8. [43] Anderson RM. Planning for pandemics of infectious diseases. The Bridge. 2006; 36(3): 5–9. [44] Clancy C, Callaghan M, Kelly T. A multi-scale problem arising in a model of avian flu virus in a seabird colony. Journal of Physics: Conference Series. 2006; 55: 45–54. [45] Heuer AH. Engineering and vaccine production for an influenza pandemic. The Bridge. 2006; 36(3): 3–4. [46] Institute of Medicine. Financing Vaccines in the 21st Century: Assuring Access and Availability. Washington D.C.: National Academies Press; 2003. [47] McCluskey MM, Alexander SB, Larkin BD, Murguia M, Wakefield S. An HIV vaccine: as we build it, will they come? Health Affairs (Project Hope). 2005 May–Jun; 24(3): 643–51. [48] Vaccine Vial Monitors (VVMs). Seattle: PATH; 2005. [49] Safe Vaccine Handling, Cold Chain and Immunizations. Geneva: World Health Organization; 1998. [50] HealthTech Historical Profile: Technologies for Injection Safety. Seattle: PATH; 2006 January.

[51] BD Immunization. 2007 [cited 2007 May 1]; Available from: immunization.asp [52] Levine MM. Can needle-free administration of vaccines become the norm in global immunization? Nature Medicine. 2003 Jan; 9(1): 99–103. [53] GAVI Alliance Progress Report. 2007; Available from: publications/index.php [54] Beal J, Orrick J, Alfonso K, Rathore M, eds. HIV/AIDS Primary Care Guide: Florida/Caribbean AIDS Education Training Center; 2006. [55] Centers for Disease Control and Prevention. Frequently Asked Questions About HIV and HIV Testing [cited 2007 May 3]; Available from: subindex.cfm?FuseAction=FAQ#1 [56] Chou R, Huffman LH, Fu R, Smits AK, Korthuis PT. Screening for HIV: a review of the evidence for the U.S. Preventive Services Task Force. Annals of Internal Medicine. 2005 Jul 5; 143(1): 55–73. [57] Smith KA. The HIV vaccine saga. Medical Immunology. 2003 Feb 14; 2(1): 1. [58] Baltimore D, Heilman C. HIV vaccines: prospects and challenges. Scientific American. 1998 July; 279(1): 98–103. [59] Letvin NL. Progress and obstacles in the development of an AIDS vaccine. Nature Reviews. 2006 Dec; 6(12): 930–9. [60] Nowak MA, McMichael AJ. How HIV defeats the immune system. HIV: 20 Years of Research: Scientific American; 2003. [61] Carmichael M. How it began: HIV before the age of AIDS. PBS Frontline. 2006 May 30. [62] Ezzel C. Hope in a Vial. HIV: 20 Years of Research: Scientific American 2003: 38–43. [63] Singh M. No vaccine against HIV yet – are we not perfectly equipped? Virology Journal. 2006; 3: 60. [64] Eaton L. AIDS vaccine may offer hope only for some ethnic groups. BMJ (Clinical research edn. 2003 Mar 1; 326(7387): 463. [65] AIDS vaccine only limited success. BBC News. 2003 February 24. [66] Kresge KJ. VRC starts Phase II vaccine trial. VAX. 2005 October; 3(10).

9 Ethics of clinical research

The practice of medicine cannot improve in the absence of medical research. Advancing clinical medicine requires controlled experiments to compare the performance of a new intervention to the current standard of care. In many cases, initial experiments can be carried out in the laboratory using cell cultures or animal models, but eventually new techniques must be tested in humans to ensure that they are safe and effective. Unfortunately as we will see, people have not always treated each other humanely in the pursuit of medical research. How do we ensure that medical research involving human subjects is carried out in a fair and ethical manner? In Chapter 9, we will examine the ethical principles that guide research involving human subjects, and how we ensure that researchers adhere to these principles. For centuries, the actions of physicians have been guided by the Hippocratic principle of “first do no harm.” This principle guides clinical practice to improve an individual patient’s health. Often the goal of medical research is to improve the health of future patients, and a subject participating in a research project may receive absolutely no benefit. In fact, participating in a research study may involve risks not fully understood at the beginning of a study. In the 1800s, scientists began to formally articulate ethical principles to guide medical research. In his 1865 book, An Introduction to the Study

of Experimental Medicine, Claude Bernard stated that one could never perform an experiment on man “which might be harmful to him in any extent, even though the result might be highly advantageous to science” [1]. For many years, ensuring that scientists and physicians adhered to these ethical principles was largely left to the discretion of individual researchers, not always with success. Table 9.1 chronicles some historical examples of ethically questionable research involving human subjects. The atrocities committed by the Nazis and by Japanese forces in World War II in the name of medical research shocked the world, and led to a new era in the regulation of medical research [15]. As a result, several codes governing the ethical conduct of research have been developed to provide guidelines for patients, practitioners and scientists. Later in Chapter 9, we will examine the ethical principles laid out in The Nuremberg Code of 1949, The Helsinki Declaration of 1964, and The Belmont Report of 1979. But we begin Chapter 9 by examining several case studies of research carried out in the United Sates which further motivated the development of these codes of conduct.

Tuskegee Syphilis Study The Tuskegee Syphilis Study was begun in 1932 in Macon County, Alabama. The goal of this study was


Biomedical Engineering for Global Health

Table 9.1. Historical examples of ethically questionable research [2–14]. Year



Edward Jenner injects healthy eight year old James Phipps with cowpox, then six weeks later with smallpox. Ultimately Jenner’s experiments gave rise to the first smallpox vaccine [2].


J. Marion Sims performs experimental surgeries on enslaved African women in an attempt to repair vesicovaginal fistulas – a severe complication of prolonged childbirth. While historical record suggests that the women voluntarily participated, Sims has been criticized for experimenting on a vulnerable population [3].


Dr. Arthur Wentworth performs spinal taps on 29 infants and children at Children’s Hospital in Boston to determine if the procedure is harmful. Upon reporting results, Wentworth is criticized by peers for failing to obtain parental consent and for performing non-therapeutic procedures [4].


Italian bacteriologist Giuseppe Sanarelli injects five subjects with what he believes to be a filtered, inactivated solution of the yellow fever bacillus, producing yellow fever like symptoms in several of the patients [5]. The experiment was carried out without the subjects’ permission or consent [6]. Walter Reed and James Carroll later disprove Sanarelli, demonstrating that the injected bacillus was actually a member of the hog cholera family [7].


Richard Strong, head of the Philippine Biological Laboratory innoculates 24 inmates of a Manila prison with a cholera vaccine that is contaminated with plague. Thirteen of the inmates die [8]. It is unclear whether or not contamination was accidental [9].


Dr. Shiro Ishii, a physician and officer in the Japanese army, directs programs throughout China dedicated to biological warfare research, including the infamous Unit 731. Prisoners of Chinese and Russian nationality were innoculated with a variety of diseases including plague, typhoid, cholera, smallpox, and hemorrhagic fever. Additional experiments were carried out on local populations by contaminating wells and food sources. Precise estimates of casualties are not possible, but number likely in the thousands [10].


Nazi physicians conduct sterilization experiments on prisoners at Auschwitz and Ravensbrueck concentration camps in an effort to identify a means of carrying out mass sterilization campaigns [11].


Nazi physicians conduct typhus experiments on prisoners of Buchenwald and Natzweiler concentration camps. Prisoners were given experimental vaccines and chemical substances and infected with typhus leading to hundreds of deaths [11].


Nazi physicians conduct hypothermia experiments on approximately 300 male prisoners in Dachau concentration camp by immersing the prisoners in tanks of ice water [12].


US Chemical warfare service conducts mustard gas experiments on approximately 4000 servicemen. Soldiers were placed in gas chambers and field testing situations in order to test experimental protective clothing and collect data on exposure levels that produce injury [10].


Four hundred prisoners in the Illinois Statesville Penitentiary volunteer to participate in malaria experiments headed by Dr. Alf Alving through the University of Chicago Medical School. At the conclusion of the two year program a considerable portion of the prisoners received parole in return for their participation [9].


The US Atomic Energy Commission and Quaker Oats Company sponsor researchers at Harvard and MIT to conduct a study of nutrient absorption at the Fernald School – a residential institution for mentally disabled children. Children were fed cereals containing radioactive tracers and received calcium tracer injections. Parents, while told of a study, were not informed of the details [10].


In a series of studies, 100 young, predominantly minority boys with a personal or family history of aggression are administered fenfluramine in an effort to test whether aggression can be predicted by chemical changes in the brain. Fenfluramine has since been taken off the market due to evidence that long term use may give rise to heart valve defects in adults [13,14].

Ethics of clinical research


to examine the natural history of untreated syphilis. At the time the study began, the standard medical therapy for syphilis was to give patients heavy metals, like

twice as high as for treated patients, yet treatment was still withheld. In the 1940s even when penicillin became the clear drug of choice to treat syphilis, the study was

bismuth and arsenic. The cure rate for this treatment was less than 30%, and the side effects were sometimes fatal [16]. While this treatment did appear to reduce mortality, it was unclear whether some of the complications of syphilis were associated with the disease itself or were side effects associated with the heavy metals [17]. Because these side effects were so debilitating, the investigators felt that the treatment was potentially as

still not interrupted and the men were not informed that penicillin was available. The study continued until 1972, when a researcher voiced concern to a reporter and the study was widely reported in the media [17]. As a result of the publicity, the study ended in 1972, and participants were offered monetary reparations. In 1973, Congressional investigations into the study commenced, and the NAACP won a $9 million settlement on behalf

toxic as the disease. In an attempt to separate the side effects of treatment from the natural progression of disease, researchers recruited a group of 600 low income black men, 399 with syphilis and 201 without syphilis [16]. The researchers withheld treatment from the group

of the participants [18]. On May 16, 1997, US President Bill Clinton apologized to the surviving participants of the Tuskegee Syphilis Study [19].

with disease; they felt they could justify withholding treatment because the side effects of the treatment were potentially as serious as the symptoms of syphilis. However, the participants did not voluntarily consent to participate in a research study. In fact, they were lured to participate in the study when researchers offered free treatment for “bad blood” – a generic term then used to describe a range of symptoms. The men were misinformed that some study procedures, like spinal taps, were free “extra treatment” (Figure 9.1) [17]. Ten years after the study began, the investigators noted that the death rate of non-treated patients was

Willowbrook School Study Another ethically questionable study was the Willowbrook School Study, which was carried out from 1963 to 1966 and sought to examine the natural history of infectious hepatitis A. The study subjects were children at the Willowbrook State School, an institution for “mentally defective persons.” Subjects in the study were deliberately infected with hepatitis A by feeding them stool from infected persons. Later in the study, as the virus became better defined, subjects were injected with the virus. The investigators justified their actions because the vast majority of children admitted to the Willowbrook State School acquired hepatitis anyway. Parents of children participating in the study gave consent for their children to participate. However, during the time of this study the Willowbrook State School was at times closed to new patients due to crowding. Because the hepatitis project had its own space, in some cases the only way to gain admission to the school was to agree to participate in the study [17].

Jewish Chronic Disease Hospital Study In 1963, in the Jewish Chronic Disease Hospital Study live cancer cells were injected into debilitated patients in a hospital for the elderly. The purpose of the study was Figure 9.1. A subject in the Tuskegee Syphilis Study undergoes a lumbar spinal tap. National Archives.

to develop information about the transplant rejection process and to study rejection of cancer cells. Patients


Biomedical Engineering for Global Health

hospitalized with various chronic debilitating diseases were injected with live cancer cells. Consent to participate in the study was negotiated orally, but not documented. Patients were not told that cancer cells would be injected because researchers felt that this might scare them unnecessarily. The investigators justified this because they were reasonably certain the cancer cells would be rejected. Researchers knew that healthy patients reject cancer cell implants quickly, while cancer patients reject the same cancer cell implants much more slowly. They wanted to understand whether this was due to impaired immunity because of the cancer or a more general manifestation of debility in cancer patients [17].

San Antonio Contraceptive Study The goal of the San Antonio Contraceptive Study was to understand which side effects of oral contraceptive pills (OCPs) are due to the drug and which are simply by-products of everyday life. The study, carried out in the 1970s, was a randomized trial comparing a placebo and OCPs. Study subjects were 76 impoverished Mexican-American women with previous multiple pregnancies who had come to a public clinic seeking contraceptive assistance. The experiment was designed as a randomized, double-blind, placebo controlled trial – meaning that a fraction of the participants received placebo while the remainder received OCPs. The study utilized a cross-over design – during the middle of the trial, the placebo group was given OCPs and the OCP group was given placebo. All women were instructed to use vaginal cream as contraceptive during the study, but none of the women were told that the study involved a placebo. During the study, 11 women became pregnant, 10 while using placebo [17].

berg Code was adopted in 1949. The Nuremberg Code states that in research, voluntary consent of the human subject is absolutely essential, and the subject should be at liberty to end the experiment at any time. All research involving human subjects should yield fruitful results for the good of society, which are obtainable in no other way. Experiments involving human subjects should avoid all unnecessary mental and physical suffering, and no experiment should be performed if it is believed that death or disabling injury may occur. The degree of risk to human subjects should never exceed the humanitarian importance of the problem to be solved. Finally, research involving human subjects should be conducted only by scientifically qualified persons [11]. In an international move to establish common ethical principles to guide medical research, the World Medical Association worked to develop and adopt the Declaration of Helsinki in 1964. The primary principle established in this document is to place the interests of the individual patient before those of society, stating that the primary goal of a physician is to “protect the life, health, privacy and dignity of the human subject [20].” The Helsinki Declaration affirms many of the principles of the Nuremberg code: that research subjects must be informed of the risks of a study and must voluntarily consent to participate, even if they are minors; and that studies should be designed and conducted by scientifically qualified personnel; and that risks of a study should not outweigh possible benefits. The Helsinki Declaration calls for formal review of research protocols by independent committees. Despite these guidelines, abuses continued. Largely

As a result of these and other examples, scientists and policy makers have developed codes to govern research

as a result of publicity associated with the Tuskegee trials, the US Department of Health, Education and Welfare issued The Belmont Report (Figure 9.2), a statement of basic ethical principles and guidelines to resolve ethical problems associated with conduct of research with human subjects, in 1979 [21]. The Belmont Report drew distinctions between clinical practice and research. Clinical practice includes interven-

involving human subjects. As a result of atrocities discovered in German concentration camps, The Nurem-

tions designed solely to enhance well being of an individual patient that have a reasonable expectation of

Codes of conduct for human subjects research

Ethics of clinical research


be treated as autonomous agents, and that the selection of research subjects must be scrutinized to determine whether some participants are being selected because of easy availability, compromised position or manipulability. The Belmont Report provided guidelines for researchers to follow in order to ensure that these three principles were applied. First, researchers must obtain

Figure 9.2. The Belmont Report, published in 1979, is a statement of basic ethical principles that must be followed in research on human subjects.

success. In contrast, research involves an activity to test a hypothesis that will permit conclusions to be drawn, and will contribute to generalizable knowledge. Research should be described in a formal protocol that sets forth an objective and procedures to reach that objective. The Belmont Report established three basic ethical principles which must be followed in all research involv-

voluntary informed consent from all study participants. In order for a participant to give informed consent, they must fully understand the research procedure, the purpose of study, the potential risks and anticipated benefits, any alternative procedures that are available to them and they must be told that they may withdraw from the study at any time. Researchers must present this information in a way the subject can understand. It cannot be disorganized, presented too rapidly, or be above the subject’s educational level. This consent must be given voluntarily, and persons in positions of authority cannot urge a particular course of action [21]. Second, research must be justified based on a favorable risk/benefit ratio for the participants, and

Respect for persons. Respect for persons demands

researchers must select subjects fairly. Here, risk is defined as the possibility that harm may occur and benefit is defined as a positive outcome related to the health or welfare of a participant. Brutal or inhumane

that subjects enter into research voluntarily with enough information to make a decision about

treatment of subjects is never justified. Instead, studies should be designed to reduce risks to only those nec-

whether to participate. Further, persons with diminished autonomy (e.g. prisoners, children) are entitled to special protection. Beneficence. Beneficence requires that researchers design experiments which do not harm study participants. Experiments which will injure one person are not allowed regardless of benefits that may

essary to achieve the research objective. Researchers must also select subjects fairly. They must not select only “undesirable” persons for risky research. Distinctions should be drawn between groups that should and should not be asked to participate in research based on ability of that group to bear burdens. For example, adults should be asked to bear burdens of research

come to others. Instead, researchers must make every effort to secure the well being of study participants, by maximizing all possible benefits and

before children, when possible. Methods used to avoid exploiting vulnerable patients include: choosing subjects who are not vulnerable, distributing benefits so

minimizing all possible harms. Justice. This principle addresses who should receive benefits of research and who should bear its burdens. Justice requires that all individuals should

that those who participate benefit, getting community consultation to hear many points of view from those being studied, and using lottery systems when there are insufficient pools of new therapy [21].

ing human subjects [21].


Biomedical Engineering for Global Health

A summary of the history of regulations Fifth Century BC: Hippocratic Oath The medical ethics standard “first do no harm” is attributed to Hippocrates. The oath became mandatory for physicians prior to practicing medicine in the fourth century AD [22]. 1949: Nuremberg Code Nazi physicians were charged with war crimes for research atrocities performed on prisoners of war. An American military war crimes tribunal conducted the proceedings against 23 Nazi physicians and administrators who willingly participated in war crimes. The judgment, known as the Nuremberg Code, was the first internationally recognized code of research ethics. It set forth ten standards for human subject research [11]. Volunteers must freely consent to participate in research. Researchers must fully inform volunteers concerning the study. Risks associated with the study must be reduced where possible. Researchers are responsible for protecting participants against harms. Participants can withdraw from the study at any time. Research must be carried out by qualified researchers. If adverse effects emerge, research must be stopped. Society should benefit from study findings. Research on humans should be based on previous animal or other work. No research study should begin if there is a reason to believe that death or injury may result. 1964: Helsinki Declaration The 18th World Medical Assembly met in Helsinki, Finland, and issued recommendations to guide biomedical research involving human subjects. The primary principle of the Declaration of Helsinki was to place individual patient interests before those of society. The basic principles of the Declaration of Helsinki are as follows [20]. The physician’s duty is to protect the life, health, privacy and dignity of the human subject. Research involving humans must conform to scientific principles and methods. Research protocols should be reviewed by an independent committee. Research protocols should be carried out by scientifically and medically qualified individuals. The risks and burden to human subjects should not outweigh the benefits. Research should be stopped if risks are found to outweigh potential benefits. Research is justified only if there is a reasonable likelihood that the population under study will benefit from the results. Participants must be volunteers and informed about the research study. Every precaution must be taken to respect privacy, confidentiality, and participants’ integrity. Consent must be obtained from minors if they are able to do so. Investigators are obliged to preserve the accuracy of results; negative and positive results should be publicly available. 1979: The Belmont Report National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research published The Belmont Report which set forth three basic ethical principles to guide research involving human subjects [21]. Respect for persons: participants must give voluntary consent; participants with diminished autonomy (e.g. children, prisoners) are entitled to special protection. Beneficence: research must maximize possible benefits and minimize possible harms. Justice: the benefits and risks of research must be distributed fairly.

Ethics of clinical research Reexamining the Tuskegee study in light of the three principles of The Belmont Report illustrates its many ethical failures. Participants did not give consent to participate, and they were not informed of the study. Risks to participants were not minimized; indeed, participation increased risks. Participants were limited to disadvantaged, rural black men, but the disease under study is not limited to this population. A much broader population benefited from the findings of the research. How do institutions work to ensure that studies conform to these guidelines? Today, US institutions carrying out research involving human subjects have a special, independent committee called the Institutional Review Board (IRB). The role of the IRB is to work with investigators to be sure that the rights of subjects are protected, to educate the research commu-


Conceptual framework for the process of obtaining informed consent

Information provision and sharing by the research team with the participants and community leaders (communal assent and agreement with the family/community) Discussion and interaction between researchers and potential participants

True understanding Acceptance or rejection of participation Agreement to participate (written, verbal, witnessed or recorded)

End of contact


Figure 9.3. The steps involved in obtaining informed consent. Used with permission from [15].

nity and public about ethical conduct of research, and to be a resource center for information about Federal guidelines. Research involving human subjects cannot begin until the IRB has approved the research protocol and the informed consent document, a written document that subjects sign indicating their willingness to participate. An IRB approved informed consent document can be found in the appendix to this chapter. The research protocol is written for review by the physicians and scientists who are members of the IRB, while the informed consent document is written for potential participants. Informed consent is a critically important part of research. The Nuremberg Code speaks to the voluntary consent of human subjects being essential: “This means the person involved should have ‘legal capacity’ to give consent; should be situated to exercise ‘free power of choice’, without the intervention of any element of force, fraud, deceit, duress, over-reaching or other ulterior motives; over-reaching or other ulterior form of constraint or coercion, and should have sufficient ‘knowledge’, and ‘comprehension’ of the elements of the subject matter involved as to enable him to make an understanding and enlightened decision [11].” Therefore, for informed consent to be valid: the subject must be competent, the consent voluntary, their participation informed, and their understanding complete.

Figure 9.3 provides an overview of the process of obtaining informed consent. The research team must provide full and understandable information about the proposed research. The participant must understand what is being asked of him or her and must freely agree to participate. Comprehension is a key element in the informed consent process; the investigator must ensure that the subject understands both the risks and benefits involved in participation. Technical procedures must be explained in lay terms at the appropriate educational level and using interpreters and translators as necessary. Researchers must document that participants have given informed consent. Most frequently, consent is documented by having participants sign a written informed consent document. Table 9.2 shows the elements typically included in such a document. The appendix to this chapter provides a sample informed consent document; as you will see, informed consent documents often use complex language and seem to be written to provide legal protection to researchers rather than to provide information for participants [15]. Unfortunately, there is currently little emphasis on assessing a participant’s understanding of a project before they sign an informed consent document. Researchers are not required to test or document participant understanding,


Biomedical Engineering for Global Health

Table 9.2. The components of an informed consent document for human research subjects. Invitation

Clear invitation to participate

Statement of overall purpose

Explanation of the purpose of the research in laymen’s language

Basis for selections

Why have you, the individual patient, been asked to participate in this study

Explanations of procedures

A description of procedures to be followed, with identification of any procedures that are experimental. A statement of where and when the research will be done, and how much time will be involved in participating in the research

Description of the discomforts and risks

Description of foreseeable risks, discomforts, and inconveniences to the subject, the likelihood that they may occur, and steps taken to minimize risk

In case of injury

Description of the availability of medical therapy as well as the compensation for disability that may result from participating

Description of benefits

Description of benefits are hoped for but not guaranteed. If participants will not benefit, this must is to develop knowledge useful in developing improved be indicated, e.g. “the purpose of therapies for your disease. Thus we hope to provide benefits in the future for persons like you”

Disclosure of alternatives

Description of alternative and routine therapies

Confidentiality assurances

Disclosure of who may review the chart; this usually involves discussion of who is supporting the study, who monitors trials for that group, and any other state or federal authorities likely to review the research work

Financial considerations

Description of any economic advantages in participating in a clinical trial, such as any financial inducements for participation, and explains that patients are usually not eligible for patent or royalty rights of invention

Offer to answer questions

Information about how to contact scientific, medical and administrative personnel in case the participant has questions regarding the study.

Continuing disclosure

Statement that the PI will notify subjects of any new findings obtained during the course of the study that may impact their decision to continue to participate in the research

although it has been suggested that simple questionnaires or interviews could be used to document understanding prior to informed consent (Figure 9.4) [15].

Continuing controversies Despite explicit ethical guidelines, recently, a number of high profile ethical dilemmas have arisen in research projects involving human subjects. We conclude Chapter 9 by reviewing the debate surrounding some of these dilemmas.

Blinded seroprevalence studies In 1988, the CDC and state health departments carried out studies to determine the prevalence of HIV in

Figure 9.4 The use of street theater can improve community c Giacomo Pirozzi/Panos knowledge to facilitate informed consent.  Pictures.

the population. They tested blood samples for HIV to determine the portion of the population infected with HIV. The study was blinded, so that researchers did not

have access to any patient identifiers. Before proceeding with the research, it was reviewed for ethical concerns.

Ethics of clinical research Informed consent was considered unnecessary because the data had been anonymized, and the researchers did not have access to information which could identify the subjects. However, this prevented the researchers from notifying infected individuals. As treatments evolved for HIV, and the importance of early clinical intervention with anti-retroviral drugs was revealed, the studies came under attack. Several legislators argued that the studies should be unblinded. Nettie Mayersohn, a democratic representative in the New York State Assembly expressed concern that infected babies who were identified through the study had a right to treatment if their test results were positive [23]. US Congressman Gary Ackerman introduced legislation to unblind the study. Ackerman warned, “There was one point in our society, a very dark day when people were allowed to walk around after being tested with a dread disease just so the medical establishment could . . . see what happens . . . ” [24]. Because of these concerns, the CDC suspended the study in 1995 [23]. Did this study adequately protect the rights of human subjects? Most experts agree that the study conformed to the ethical guidelines of The Belmont Report. These guidelines permit experiments to be carried out using patient specimens which will normally be discarded without consent, so long as patient identities are not released to investigators and the study has been reviewed by an IRB. The purpose of the study was to identify populations at risk for HIV so that effective interventions could be designed for these groups. None of the study participants were prevented or discouraged from seeking voluntary HIV testing [23].

Study of HIV transmission in Uganda


antibiotics lowered the rate of other STDs, but did not affect the rate of HIV transmission. When the study was ended all participants were given antibiotics. After the study was concluded, the researchers analyzed their data to see what other factors might affect HIV transmission. They matched sexual partners and identified 415 partners where one partner was infected and the other was not at the beginning of the study. They found that the most significant factor likely to increase transmission from the infected to the uninfected partner was the amount of virus in the infected person’s blood [26]. The study was criticized by Marcia Angell, editor of the New England Journal of Medicine, who was troubled that the researchers did not inform the at-risk partners. The researchers did not identify the discordant couples until after the study had been concluded, and argued that even if they had known, they could not have informed at risk partners, because Uganda has a national policy that prevents health workers from telling a third party about an individual’s HIV status [26]. Angell was also troubled that HIV positive participants were not offered treatment with anti-retroviral drugs. Angell believed that the Helsinki Declaration requires that researchers provide the best available treatment to their subjects. She argued that it did not matter that such care is not usually available in the setting where the research was conducted; the researchers had an ethical obligation to provide the same treatment that would be available in a developed country [26]. Edward Mbidde, a medical oncologist in Uganda, pointed out that if studies in the developing world were held to the same standards of care available in developed countries, research to develop new treatments affordable for use in developing countries would be impractical [26].

From 1994 to 1998 a team led by researchers at Columbia University tested 15,000 adults in ten rural Ugandan communities for HIV and other sexually trans-

Developing country HIV prevention trials

mitted diseases (STDs) [25]. The goal of the study was to determine whether treatment of STDs like syphilis and chlamydia could reduce the transmission of HIV. All participants in five villages were treated for STDs, while participants in five control villages were simply told of their results and were referred to free clinics for treatment. Results showed that treatment with

In the USA, a study was carried out to determine whether treatment could interrupt transmission of HIV from mothers to babies. The trial was called the AIDS Clinical Trial Group (ACTG) Study 076 [27]. It showed a dramatic reduction in transmission for women who received the intervention compared to women who received placebo. The effect was so dramatic that the


Biomedical Engineering for Global Health

study was stopped early, so that no additional women received placebo. In this trial, drug was administered during the last 26 weeks of pregnancy. Drug was also given intravenously during delivery and to the baby for six weeks after delivery. While successful, the intervention cost $800 for drug alone [28]. Because of this high cost and the long duration over which drug must be given, many people believed its use would be impractical in many developing countries, where women don’t deliver in hospitals, don’t seek care until later in their pregnancies, and can’t afford an $800 drug [28]. Studies were started in nine developing countries to determine whether much cheaper alternatives could also reduce maternal to child transmission of HIV. The goal of these studies was to evaluate the effectiveness of a regimen which provided drug only during the last

Bouncing PORECO Babies!: June 19, 2007 Dave Swaziland Today is the one-year anniversary of the initiation of these babies into the clinic’s PORECO program. [PORECO stands for Pilot Operational Research and Community Based Project.] The aim of the PORECO program is to prevent the transmission of HIV from mothers to their babies. To celebrate, we have a huge, bouncy, inflatable castle . . . and an enormous, enormous cake (like four feet by five feet)! And, boy, are they happy today!

three to four weeks of pregnancy, reducing the cost of the intervention to just $80 [29]. This could be afforded by two of the countries, and international agencies made a commitment to provide drug to other resource poor countries participating in the trials [28]. The trial was designed as a randomized trial in which some mothers got the new regime and others received a placebo. These trials were sponsored by the CDC and the NIH and all were subject to careful ethical review [28]. The study led to a bitter ethical debate regarding the appropriate standard of care to be used in the control arm. Marica Angell, editor of the New England Journal of Medicine, criticized the trials on September 18, 1997, saying “The justifications are reminiscent of the Tuskegee study: Women in the Third World would not receive antiretroviral treatment anyway, so the investigators are simply observing what would happen to the subject’s infants if there were no study [30].” She cited the Declaration of Helsinki as preventing the trials. Angell argued that the new intervention should have been compared to the full ACTG 076 protocol, which was the standard of care in the developed world. Researchers argued that investigators would learn more in a shorter time if they did a placebo controlled trial. The placebo control was necessary to establish the baseline rates of maternal to child HIV transmission, because these vary throughout the world. Rates

of transmission can be influenced by the health state of the mothers and babies. Mothers in developing countries are often anemic and malnourished, so researchers wanted to measure the baseline transmission rate in order to know whether the new treatment reduces the rate of transmission below the baseline rate. Also, the drug itself causes anemia, so researchers believed that a placebo control was needed to determine whether the drug increased anemia. Other ethicists argued that the trial was ethical only if it was accompanied by a plan to make the treatment available to the local population if it proved to be effective [29]. Amidst the controversy, the CDC sponsored study in Thailand took place and showed that the reduced course of therapy did dramatically reduce maternal to child HIV transmission rates – although not as much as

Ethics of clinical research


Most experts agree that a new definition of the standard of care is needed, which permits different standards for research in developing countries. However, these discrepancies should be subject to approval by ethical review committees in the host country. Rather than requiring that patients have access to the highest attainable standard of care, it has been suggested that researchers provide access to highest attainable and sustainable therapeutic method. The level of therapy gen-

Figure 9.5. A billboard in Gabarone, Botswana advocates participation in programs to Prevent Maternal to Child Transmission (PMTCT).

the ACTG 076 protocol. Within weeks after the study findings were made public, agencies started supplying drug to women in studies around the world who were previously on placebo. Glaxo Wellcome, the drug manufacturer announced it would cut prices of drug for sale in developing countries. Thus, the study enabled worldwide programs designed to prevent maternal to child transmission of HIV (Figure 9.5) [29].

Standard of care: a new definition? Many of the current controversies center on debate over what should be the appropriate standard of care for research involving human subjects. How do we decide what is a reasonable standard of care for research subjects in developing countries? Should we automatically use the standard set by developed countries? What is the danger of simply imposing the highest attainable standard of care for all research throughout the world? If we require that subjects in the control group receive the same treatment that would be available to them in a developed country, we may never develop sustainable techniques to improve health in developing countries [28]. What is the danger of accepting less than the highest attainable standard of care? We may find that researchers choose to carry out phase I drug studies in Africa because it is cheaper and less regulated.

erally available in a host country is the least that is ethically acceptable. Researchers must commit to provide a level of treatment that can continue in a host country after the research has been completed. If therapy is not

Suggested guidelines for research involving human subjects in developing countries Carry out research on a health problem of the developing country population. Research objectives, not vulnerability of the population, should be used to justify conduct of the research in a developing country. Ensure that benefits of participating in trial outweigh the risks. Only undertake research that benefits the community participating. Translate research findings into accessible care in the community participating. Involve members of the host community in design and conduct of trial; they must decide if benefits outweigh risks. Provide subjects with care they would not ordinarily get in the country where the trial is carried out. Ensure that trial does not widen disparities by taking resources away from the healthcare system of the host country. Interventions proven safe and effective through research should be made available in those countries [28, 31, 32].


Biomedical Engineering for Global Health

sustainable, then results can never be made available to the inhabitants of the country [28].

Bioengineering and Global Health Project Project task 5: Define the constraints that a solution must satisfy These should be quantitative measures that include both technical performance and economic constraints that your solution must satisfy. If there are existing solutions, you should identify the performance capabilities and cost of these solutions. Your solution should provide an advantage compared to existing solutions. You should carefully justify trade-offs made between expected performance and cost. Examples of constraints that

Blood trial could omit consent form Doctors seek community consensus to test a blood substitute on trauma patients who may not be conscious ANDY DWORKIN How would you feel knowing that a doctor could experiment on you, without your permission, while you were unconscious? What if that experiment could help save your life and test a possible treatment for wounded soldiers or car crash victims? Doctors want Portland-area residents to ponder those questions as they move toward joining a study of a blood substitute called PolyHeme. Trauma medics with Legacy Health System, Oregon Health & Science

you might consider include necessary educational level of primary user, detection limits of new diagnostic methods, efficacy rates of new therapies, power requirements, cost and size. Turn in a one page table summarizing the design constraints for

University and local ambulance companies would take part in a national trial comparing PolyHeme with the salt-water solution now carried on ambulances. This is no ordinary research project. In

your problem. Each row in the table should indicate

most trials, scientists must tell each potential participant about the possible risks and rewards

a specific constraint (e.g. unit cost of device). The table should include at least two columns – one or more which represents the current performance of available technologies and one which represents the constraint that your design must satisfy.

Homework 1. The Belmont Report establishes the three fundamental ethical principles that guide the ethical conduct of research involving human participants: (1) Respect for persons; (2) Justice; and (3) Beneficence. These principles require that all subjects participating in medical research give informed consent. a. Define informed consent. b. The following story appeared in The Oregonian. Read it and answer the following question. Suppose you are a member of the OHSU IRB. Would you have voted to approve this trial? Why or why not? Support your answer using the principles of The Belmont Report.

before getting their agreement to participate, a process called “informed consent.” But PolyHeme would go to people unconscious from blood loss when treatment starts. A seldom-used and ethically controversial 1996 Food and Drug Administration regulation lets researchers waive informed consent to test potential life-saving treatments when there is no other way to conduct the research. Instead of individual consent, the FDA says researchers must teach local residents about the trial and gauge their feelings. So Legacy and OHSU workers are mailing letters to local officials and holding three public meetings to explain the trial and ask for feedback. “This is not a sure thing that the study will happen,” said Lise Harwin, a Legacy communications coordinator who helped plan the public education. “What we’re trying to do now is get feedback to determine if it will.” Portland researchers have spent more than a year planning the trial, and both hospitals’

Ethics of clinical research researchreview boards have approved the idea. Used with permission from Oregonian Pub Co., from Oregonian, Andy Dworkin, 2005; permission conveyed through Copyright Clearance Center, Inc. But those boards won’t give their final approval until they consider public reaction. Scientists have spent decades searching for a blood substitute, which trauma doctors say is desperately needed. Donated blood is too delicate and has too short a shelf life to carry on ambulances. Instead, paramedics use durable saline solution. But saline can’t carry oxygen through the body; PolyHeme does. PolyHeme, which is made from expired blood donations, has a longer shelf life than blood and can be


basic protection afforded in federal law and required by some states, an Associated Press review has found. The research funded by the National Institutes of Health was most widespread in the 1990s as foster care agencies sought treatments for their HIV-infected children that weren’t yet available in the marketplace. The practice ensured that foster children – mostly poor or minority – received care from world-class researchers at government expense, slowing their rate of death and extending their lives. But it also exposed a vulnerable population to the risks of medical research and drugs that were known to have serious side effects in adults and for which the safety for children was unknown. Several studies that enlisted foster children reported that patients suffered side effects such as

administered to a person of any blood type. Local research boards “haven’t established a particular percent or number” of negative responses from

rashes, vomiting and sharp drops in infectionfighting blood cells as they tested antiretroviral drugs to suppress AIDS or other medicines to treat

the community that would cause them to stop the trial, Allee said. One reason is that researchers assume people worried about the process are more likely to comment than those

secondary infections. In one study, researchers reported a “disturbing” higher death rate among children who took higher doses of a drug. That study was unable to determine a safe and effective

who support it.

dosage. Research and foster agencies declined to make foster parents or children in the drug trials

2. The following text contains a portion of an article which appeared in the Austin American Statesman. Read the text and answer the following questions. Federal researchers tested AIDS drugs on foster children without advocate protections At least seven states, including Texas, participated in studies, which are now under investigation. By John Solomon ASSOCIATED PRESS Thursday, May 05, 2005. Used with permission of the c 2009. All rights Associated Press Copyright  reserved. WASHINGTON – Government-funded researchers tested AIDS drugs on hundreds of foster children over the past two decades in at least seven states, including Texas, often without providing them a

available for interviews, or to provide information about individual drug dosages, side effects or deaths, citing medical privacy laws. Some foster children died during studies, but state or city agencies said they could find no records that any deaths were directly caused by experimental treatments. The government provided special protections for child wards in 1983. They required researchers and their oversight boards to appoint independent advocates for any foster child enrolled in a narrow class of studies that involved greater than minimal risk and lacked the promise of direct benefit. Some foster agencies required the protection regardless of risks and benefits. Advocates must be independent of the foster care and research agencies, have some understanding of medical issues and “act in the best


Biomedical Engineering for Global Health

interests of the child” for the entirety of the research, the law states. However, researchers and foster agencies said

have been appointed for all foster children because researchers felt the pressure of a medical crisis and knew there was great uncertainty as to how

foster children in AIDS drug trials often weren’t given such advocates even though research institutions many times promised to do so to gain access to the children. Illinois officials say they think none of their nearly 200 foster children in AIDS studies got independent monitors even though researchers signed a document guaranteeing “the appointment of an advocate for each individual

children would react to AIDS medications that were often toxic for adults. “It is exactly that set of circumstances that made it absolutely mandatory to get those kids those advocates,” Caplan said. “It is inexcusable that they wouldn’t have an advocate for each one of those children.” Those who made the decisions say the research gave foster kids access to drugs they otherwise

ward participating in the respective medical research.” New York City could find records showing 142 – less than a third – of the 465 foster children in AIDS drug trials got such monitors even though city policy required them. The city has asked

couldn’t get. And they say they protected children’s interest by explaining risks and benefits to state guardians, foster parents and the children themselves. “I understand the ethical dilemma surrounding the introduction of foster children into

an outside firm to investigate.

trials,” said Dr. Mark Kline, a pediatric AIDS expert

Researchers typically secured permission to enroll foster children through city or state agencies. They frequently exempted themselves from appointing advocates by concluding the research carried

at Baylor College of Medicine. He enrolled some Texas foster kids in his studies, and said he doesn’t recall appointing advocates for them. “To say as a group that foster children should be excluded from

minimal risk and the child would directly benefit because the drugs had already been tried in adults. If they decline to appoint advocates under the

clinical trials would have meant excluding these children from the best available therapies at the time,” he said. “From an ethical perspective, I never

federal law, researchers and their oversight boards must conclude that the experimental treatment affords the same or better risk-benefit possibilities than alternate treatments already in the market-

thought that was a stand I could take.” Illinois officials directly credit the decision to enroll HIV-positive foster kids with bringing about a decline in deaths – from 40 between 1989 and 1995

place. They also must abide by any additional protections required by state and local authorities. Many of the studies that enrolled foster children occurred after 1990 when the government approved using the drug AZT – an effective AIDS treatment – for children. Those studies often involved early Phase I and Phase II research – the riskiest – to determine side effects and safe dosages so children

to only 19 since. NIH did not track researchers to determine whether they appointed advocates. Instead, the decision was left to medical review boards made up of volunteers at each study site. A recent Institute of Medicine study concluded those Institutional Review Boards were often overwhelmed, dominated by scientists and not focused enough on patient

could begin taking adult “cocktails,” the powerful drug combinations that suppress AIDS but can cause bad reactions like rashes and organ damage. Some of those drugs were approved ultimately for children, such as stavudine and zidovudine. Others were not.


Arthur Caplan, head of medical ethics at the University of Pennsylvania, said advocates should

a. What are the three basic ethical principles of The Belmont Report? Define each principle. b. Discuss the ethical and legal issues that arise when new medical technologies are tested in vulnerable populations, such as foster children. Do you think that the studies described

Ethics of clinical research adequately protected the rights of this population? Give the reasons for your position in terms of the principles outlined in The Belmont Report. c. The article states that the studies ensured that foster children received care from world-class researchers at government expense, extending their lives. Illinois officials credit the decision to enroll HIV-positive foster children with bringing about a decline in deaths. Describe how these outcomes influence your reasoning in part b above. 3. Briefly describe the Willowbrook study to investigate the natural history of infectious hepatitis. List the principles of The Belmont Report which were violated in this study. Support your answer with evidence. 4. A clinical trial recently carried out at Johns Hopkins University tested the effects of a chemical irritant to understand why some people get asthma. Three healthy volunteers with normal respiratory systems


viral infection making the rounds at Bayview at the time. Source: “At Your Own Risk: Some Patients Join Clinical Trials Out of Desperation, Others to Help Medicine Advance,” Time, April 22, 2002. Discuss any problems associated with the protection of human subjects using the principles of The Belmont Report. 5. Use the following link to read the article “Placebos break taboo in cancer drug tests: Study seeks hope for desperately ill” that was first printed in the Boston Globe: You have just been named the Director of the National Cancer Institute. You control an annual budget of $6 billion. You must decide whether any of these funds can be used to support placebo controlled research studies for terminally ill cancer patients. Your decision will determine whether any studies of this type will receive any funding. Using the article as a reference point, prepare an argument in favor of or against such studies. Your argument

inhaled the chemical. Two days after inhaling the chemical, Ellen Roche, 24, a technician at the Johns Hopkins Asthma and Allergy Center, developed a

should be no more than one typed page. Limit your argument to either the pro or con stance and prepare a convincing case as to why you ruled the

cough, fever and muscle pain. She quickly developed respiratory distress, and within a month she was dead. The chemical she inhaled turned out to be far more toxic than the researchers realized. In

way you did. 6. Discuss the ethical and legal issues that arise when new medical technologies are tested in developing countries. In what ways can this benefit the

fact, the lead investigator’s literature search of the most common databases (which date back only to 1960), did not turn up earlier studies hinting at the chemical’s potential dangers, but after-the-fact searches using different search engines and databases did turn up references to the potential risks to humans. In a review of the study, the FDA raised questions about the informed-consent forms that Roche and two other subjects had signed. On them, hexamethonium is referred to as a “medication” and as “(having) been used as an anesthetic” – giving subjects a sense that it was an FDA-approved medicine and therefore safe. Another criticism: Togias failed to report that his first subject (Roche was the third) had developed a cough. It went away, and Togias assumed it had to do with a

population of the developing country? In what ways can the population be harmed? If the researchers are based in the United States, what legal and ethical responsibilities do they have?

References [1] Bernard C. An Introduction to the Study of Experimental Medicine. New York: Macmillan; 1927. [2] Stern AM, Markel H. The history of vaccines and immunization: familiar patterns, new challenges. Health Affairs (Project Hope). 2005 May–Jun; 24(3): 611–21. [3] Wall LL. The medical ethics of Dr J Marion Sims: a fresh look at the historical record. Journal of Medical Ethics. 2006 Jun; 32(6): 346–50. [4] Grodin MA, Glantz LH. Children as Research Subjects : Science, Ethics, and Law. New York: Oxford University Press; 1994.


Biomedical Engineering for Global Health

[5] Lederer SE. Subjected to Science : Human Experimentation in America before the Second World War. Baltimore: Johns Hopkins University Press; 1995. [6] Pierce JR, Writer J. Yellow Jack : How Yellow Fever Ravaged America and Walter Reed Discovered its Deadly Secrets. Hoboken, N.J.: John Wiley; 2005. [7] Phillip S. Hench Walter Reed Yellow Fever Collection. The United States Army Yellow Fever Commission (1900–1901). 2001 August 8 [cited 2007 May 7]; Available from: commission.html# edn1 [8] Chernin E. Richard Pearson Strong and the iatrogenic plague disaster in Bilibid Prison, Manila, 1906. Reviews of Infectious Diseases. 1989 Nov–Dec; 11(6): 996–1004. [9] Hornblum AM. Acres of Skin : Human Experiments at Holmesburg Prison : a Story of Abuse and Exploitation in the Name of Medical Science. New York: Routledge; 1998. [10] Moreno JD. Undue Risk : Secret State Experiments on Humans. New York: W.H. Freeman; 2000. [11] Trials of War Criminals before the Nuremberg Military Tribunals under Control Council Law No. 10. Nuremberg, October 1946–April 1949. Washington D.C.: U.S. G.P.O 1949–1953. [12] Michalczyk JJ. Medicine, Ethics, and the Third Reich : Historical and Contemporary Issues. Kansas City, MO: Sheed & Ward; 1994. [13] Hilts PJ. Experiments on children are reviewed. The New York Times. 1998 April 15. [14] Bernstein N. 2 Institutions faulted for tests on children. The New York Times. 1999 June 12. [15] Bhutta ZA. Beyond informed consent. Bulletin of the World Health Organization. 2004 Oct; 82(10): 771–7. [16] Centers for Disease Control. US Public Health Service Syphilis Study at Tuskegee. 2007 March 27 [cited 2007 June 1]; Available from: [17] Levine RJ. Ethics and Regulation of Clinical Research. 2nd edn. New Haven: Yale University Press; 1986. [18] Around the Nation; judge upholds deadline on syphilis case settlement. The New York Times. 1981 June 18. [19] Mitchell A. Clinton regrets ‘clearly racist’ U.S. study. The New York Times. 1997 May 17. [20] World Medical Association. Declaration of Helsinki. 52nd World Medical Association General Assembly; 2000; Edinburgh, Scotland; 2000.

[21] United States. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report : Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington: Dept. of Health, Education, and Welfare, National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research: for sale by the Supt. of Docs., US Govt. Print. Off. 1978. [22] National Library of Medicine. Greek Medicine: The Hippocratic Oath. [cited 2007 May 9]; Available from: oath.html [23] Fairchild AL, Bayer R. Uses and abuses of Tuskegee. Science (New York, NY). 1999 May 7; 284(5416): 919–21. [24] Testimony of Gary Ackerman. Hearing Before the Subcommittee on Health and Environment of the Committee on Commerce, House of Representatives. 1st Session, 11 May ed 1995:Serial No. 104–22:8. [25] Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. The New England Journal of Medicine. 2000 Mar 30; 342(13): 921–9. [26] Vogel G. Study of HIV transmission sparks ethics debate. Science (New York, NY). 2000 Apr 7; 288(5463): 22–3. [27] Connor EM, Sperling RS, Gelber R, Kiselev P, Scott G, O’Sullivan MJ, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. The New England Journal of Medicine. 1994 Nov 3; 331(18): 1173–80. [28] Levine RJ. Some recent developments in the international guidelines on the ethics of research involving human subjects. Annals of the New York Academy of Sciences. 2000 Nov; 918: 170–8. [29] Beardsley T. Coping with HIV’s ethical dilemmas. Scientific American. 1998 July; 279(1): 106–7. [30] Angell M. The ethics of clinical research in the Third World. The New England Journal of Medicine. 1997 Sep 18; 337(12): 847–9. [31] Benatar SR, Singer PA. A new look at international research ethics. BMJ (Clinical research edn). 2000 Sep 30; 321(7264): 824–6. [32] Lo B, Bayer R. Establishing ethical trials for treatment and prevention of AIDS in developing countries. BMJ (Clinical research edn). 2003 Aug 9; 327(7410): 337–9.

Ethics of clinical research


Appendix: Informed Consent Document ___________ Informed Consent to Participate in Research The University of Texas at Austin You are being asked to participate in a research study. This form provides you with information about the study. The Principal Investigator (the person in charge of this research) or his/her representative will also describe this study to you and answer all of your questions. Please read the information below and ask questions about anything you don’t understand before deciding whether or not to take part. Your participation is entirely voluntary and you can refuse to participate without penalty or loss of benefits to which you are otherwise entitled. Title of research study Evaluating the Effectiveness of Evidence-Based Teaching Strategies in BME 301: Biotechnology and World Health Principal Investigator(s) (include faculty sponsor), UT affiliation, and Telephone Number(s) Rebecca Richards-Kortum, Ph.D. Professor of Biomedical Engineering 512–471-2104 Funding source Howard Hughes Medical Institute What is the purpose of this study? The purpose of the study is to investigate the effectiveness of learner-centered, open-ended problem solving and cooperative learning strategies in BME 301. The total number of students registered for BME 301 was approximately 60. Thirty seven students participated from BME 301. Your participation will serve as a control group. What will be done if you take part in this research study? From a pool of undergraduate students, we are requesting volunteers for an interview protocol. We will ask participants to volunteer to take part in an activity in which they are given a newspaper article related to the BME 301 course material and asked to critically discuss it in groups of three to five. Participant responses will be videotaped. Students will be compensated for their time with a $20.00 gift certificate from Barnes and Noble. What are the possible discomforts and risks? There are no physical risks or discomforts that apply with this research. However, if you wish to discuss the information above or any other risks you may experience, you may ask questions now or call the Principal Investigator listed on the front page of this form. Some participants may be uncomfortable with being videotaped initially. If you feel uncomfortable, you may discontinue participation at any time. The only treatment for participating in this research that differs from discussing a newspaper article casually with a peer is your agreement to being videotaped. What are the possible benefits to you or to others? Your participation in this study will provide data which will permit researchers to identify learning behaviors that are positively associated with content mastery in this course when comparing various instructional techniques. This study will result in a deeper understanding of learner centered environments and the effect of this on learning. The study may result in the development of reliable and valid instruments which can measure learning in more effective ways than are currently used. This will improve the knowledge base for the


Biomedical Engineering for Global Health

science of learning and ultimately the knowledge disseminated from the study could improve the teaching of undergraduate curricula beyond the University of Texas at Austin. If you choose to take part in this study, will it cost you anything? There are no costs associated with participating in this study. Will you receive compensation for your participation in this study? What if you are injured because of the study? Students who complete the interview will be compensated for their time with a $20.00 gift certificate from Barnes and Noble. There are no physical risks associated with this research. University students may be treated at the usual level of care with the usual cost for services at the Student Health Center for any injury, related or not, but no payment can be provided in the event of a medical problem. If you do not want to take part in this study, what other options are available to you? Participation in this study is entirely voluntary. You are free to refuse to be in the study, and your refusal will not influence current or future relationships with The University of Texas at Austin or your grades in any course. Your decision to participate will not bestow any competitive academic or occupational advantage over any other University of Texas at Austin students who do not volunteer, and the researchers will not impose any academic or occupational penalty on those University of Texas at Austin students who do not volunteer. How can you withdraw from this research study and who should you call if you have questions? If you wish to stop your participation in this research study for any reason, you should contact: Deanna Buckley at (512)-471–3068. You are free to withdraw your consent and stop participation in this research study at any time without penalty or loss of benefits for which you may be entitled. Throughout the study, the researchers will notify you of new information that may become available and that might affect your decision to remain in the study. In addition, if you have questions about your rights as a research participant, please contact Clarke A. Burnham, Ph.D., Chair, and The University of Texas at Austin Institutional Review Board for the Protection of Human Subjects, 512–232-4383. How will your privacy and the confidentiality of your research records be protected? Authorized persons from The University of Texas at Austin and the Institutional Review Board have the legal right to review your research records and will protect the confidentiality of those records to the extent permitted by law. If the research project is sponsored, then the sponsor also has the legal right to review your research records. Otherwise, your research records will not be released without your consent unless required by law or a court order. If the results of this research are published or presented at scientific meetings, your identity will not be disclosed. Because these studies will use video recordings, you should know that the CDs will be: (a) coded so that no personally identifying information is visible on them; (b) kept in a secure locked location in the co-investigator’s office (Deanna Buckley); (c) heard or viewed only for research purposes by the investigator and his or her associates; (d) possibly retained for future research analysis. Will the researchers benefit from your participation in this study? Your participation will allow researchers to collect objective data to be analyzed for publications in educational and scientific research journals and presentations to other scientific researchers and educators. No other benefits are expected at this time.

Ethics of clinical research


Signatures As a representative of this study, I have explained the purpose, the procedures, the benefits, and the risks that are involved in this research study. ______________________________________________________________________________________________ Signature and printed name of person obtaining consent Date You have been informed about this study’s purpose, procedures, possible benefits and risks, and you have received a copy of this Form. You have been given the opportunity to ask questions before you sign, and you have been told that you can ask other questions at any time. You voluntarily agree to participate in this study. By signing this form, you are not waiving any of your legal rights. ______________________________________________________________________________________________ Printed Name of Subject Date ______________________________________________________________________________________________ Signature of Subject Date ______________________________________________________________________________________________ Signature of Principal Investigator Date We may wish to present some of the tapes from this study at scientific conventions or as demonstrations in classrooms. Please sign below if you are willing to allow us to do so with the tape of your performance. ______________________________________________________________________________________________ Signature of Subject


I hereby give permission for the video tape made for this research study to be also used for educational purposes. ______________________________________________________________________________________________ Signature of Subject


10 Technologies for early detection and prevention of cancer

In Chapter 8, we saw how technology could be used to prevent infectious diseases, one of the leading killers

prostate cancer – and look at existing and new technologies to aid in the early detection and prevention of each

in the developing world. In this chapter, we examine how technology can be used to diagnose disease. Our focus is the detection of cancer, where early detection (Figure 10.1) can mean the difference between life and


death. We begin by examining the global burden of cancer. Next, we examine how cancers develop and why early detection is so crucial. Finally, we examine three cancers in detail – cervical cancer, ovarian cancer and

The burden of cancer in the United States Cancer is the second leading cause of death in the United States, responsible for nearly one out of every four deaths (Table 10.1). Only 66% of all cancer patients live more than five years past their initial diagnosis, a statistic known as the five year survival rate. Cancer is

Figure 10.1. Mammography is one method used to screen for breast cancer. Source: Mitchell D. Schnall, M.D., Ph.D. University Of Pennsylvania, National Cancer Institute.

Technologies for early detection and prevention of cancer Table 10.1. Cancer is the second leading cause of death overall in the USA, and the leading cause of death for people under 85 years of age [2].


A comprehensive set of statistics regarding cancer incidence and mortality in the United States is compiled each year by the National Cancer Institute; the report, called Cancer Facts and Figures, can be found at It was predicted that 1,444,920 new cases of cancer would be detected in the USA in 2007, and that 559,650 people would die as a result of cancer [1]. Table 10.2 ranks the most commonly occurring cancers in men and women in the United States (excluding basal cell and squamous skin cancers). Prostate cancer is the most common cancer in US men, accounting for 1/3 of cancer incidence. Breast cancer is the most common cancer in USA women, accounting for nearly 1/3 of

important from an economic perspective as well. In the USA alone, cancer cost approximately $206 billion in 2006. Of this, $78.2 billion was spent on direct medical costs; $17.9 billion represents lost productivity to illness and $110.2 billion represents lost productivity due to premature death [1].

cancer incidence; ovarian cancer, which we will study in detail later, accounts for 3% of cancer incidence in US women [3]. Table 10.3 ranks the most common causes of cancer mortality in men and women in the United States. In both sexes, lung cancer is the leading cause of cancer

Table 10.2. The most commonly occurring cancers in men and women in the USA in 2005.

Adapted and reprinted with permission from American Cancer Society. Cancer Facts and Figures 2009. Atlanta: American Cancer Society, Inc.


Biomedical Engineering for Global Health

Table 10.3. The most common causes of cancer mortality in men and women in the USA in 2005.

Adapted and reprinted with the permission from the American Cancer Society. Cancer Facts and Figures 2009. Atlanta: American Cancer Society, Inc.

tests for lung cancer, it tends to be diagnosed at a much

Trends in cancer incidence and mortality in the USA From 1993 to 2002, cancer death rates in the United States dropped by 1.1% per year. The decrease in cancer mortality is attributed to a combination of better treatment, better early detection and cancer prevention. Death rates dropped more for men (1.5%/year than for women 0.8%/year). Cancer incidence rates in the United States have been stable since 1992 [4].

later stage with worse prognosis. The situation is similar for ovarian cancer. Although it is responsible for only 3% of cancer incidence in women, it accounts for 6% of cancer mortality in women [3].

The global burden of cancer Globally, cancer is an important cause of mortality, accounting for 12% of all deaths worldwide (Figure 10.2). Cardiovascular disease is the leading cause of mortality worldwide, followed by infectious disease and cancer. Today, more than 11 million new

death, even though it is only the second most common cancer in men and women separately. As we will see later in this chapter, routine tests are available to screen

cases of cancer are detected worldwide every year, and 6.7 million deaths can be attributed to cancer [6]. Table 10.4 shows the leading causes of cancer mortality worldwide. In men, lung cancer is the most common

older men and women for prostate cancer and breast cancer, so they tend to be diagnosed at an earlier, more curable stage. Because we do not have good screening

cause of cancer death, while in women, breast cancer is the most common cause of cancer death. The third most common cause of cancer death in women worldwide is

Technologies for early detection and prevention of cancer


Table 10.4. The number of estimated cancer deaths worldwide in 2002.

Source: International Agency for Research on Cancer (IARC). GLOBOCAN 2002 estimates.

be attributed to the use of screening tools to detect cervical cancer at an early stage. In the developed world, the Papanicoloau (Pap) smear is used to screen the general female population for cervical cancer and its precursors. The early detection and treatment of these conditions prevents the development of invasive cervical cancer. Unfortunately, owing to limited resources, cervical cancer screening is not implemented in many developing countries; as a result, cervical cancer is the leading cause of cancer death for women in developing countries [6]. The maps in Figure 10.3 illustrate global variations in the mortality of cancer today, and the changes predicted in cancer mortality throughout the world in the Figure 10.2. The most common causes of death worldwide in 2002. Cancer is the third leading cause of mortality, worldwide [5].

cervical cancer; note that cervical cancer was not among the top ten causes of cancer incidence or mortality in the USA. Again, as we will see later, this large difference can

year 2020. Both the global incidence and mortality of cancer are predicted to increase. In the next 20 years, it is estimated that global cancer incidence will increase by nearly 50% and global cancer mortality will double. The largest rates of increase are predicted to occur in developing and newly industrialized countries. In 2020,


Biomedical Engineering for Global Health



Figure 10.3. (a) Global distribution of cancer deaths in 2002. (b) Predicted global distribution of cancer deaths in 2020. Cancer is predicted to kill more than 10 million people worldwide each year.

Technologies for early detection and prevention of cancer


Table 10.5. Risk for US women of developing cancer over the course of a lifetime.

Table 10.6. Risk for US men of developing cancer over the course of a lifetime.

Slides reprinted with the permission from the American Cancer Society, Inc. from All rights reserved.

Slides reprinted with the permission from the American Cancer Society, Inc. from All rights reserved.

more than 16 million new cancer cases are predicted, and 10.3 million people are expected to die of cancer

are responsible for nearly 25% of cancers, while only 9% of cancers are due to infectious agents in devel-

[7]. Although the probability of being diagnosed with cancer is twice as high in developed countries, cancer survival rates are much lower in developing countries. In developed countries, about 50% of cancer patients

oped countries. These include hepatitis B and C (can lead to liver cancer), the sexually transmitted human papillomavirus (HPV) (can lead to cervical cancer), and Helicobacter pylori (can lead to stomach cancer). As

die as a result of their cancer, while in developing countries more than 80% of cancer patients already have

we will see later, vaccination may be the key to preventing these cancers [11]. Figure 10.4 shows the rela-

late-stage incurable tumors at the time of their diagnosis [8]. It is estimated that by 2020, at least 70% of cancer deaths will occur in developing countries, where resources for early detection and treatment are least

tionship between per capita cigarette consumption and lung cancer rates in men and women [12]. There is a 20–25 year delay between the peak in cigarette consumption and the peak in lung cancer incidence, reflect-

available [9]. If you live in the USA, what is your lifetime risk of

ing the long period of exposure and resulting biological changes which occur in lung cancer. Rates of lung can-

developing cancer? If you are female, you have a 33% chance of developing cancer at some time in your life, with a 14% chance of developing breast cancer at some point in your life (Table 10.5). If you are male, you have

cer incidence in women peak about a decade later than in men, reflecting the delay in when women began to smoke. While tobacco use has declined in many developed countries, it is rising in many developing countries.

a 50% chance of developing cancer at sometime in your life, with nearly a 17% chance of developing prostate cancer at some point (Table 10.6) [10]. How can you reduce your cancer risk? More than 1/3 of cancers are preventable, through three approaches: (1) reducing tobacco use, (2) implementing existing screening techniques worldwide, and (3) adopting a healthier lifestyle and diet. Globally, 43% of cancer deaths are due to tobacco use, inappropriate diet or infection [7]. In developing countries, infectious agents

Worldwide about 35% of men in developed countries smoke, while the fraction of men who smoke is 50% in developing countries. China represents a particular concern; with more than 300 million male smokers today, future increases in lung cancer incidence and mortality are a likely consequence [13]. Changes in diet can also reduce cancer risk. The American Cancer Society recommends that persons consume five servings of fruits and vegetables daily to reduce their cancer risk. Unfortunately, less than 1/4


Biomedical Engineering for Global Health







3500 3000


Per capita cigarette consumption

60 50

2500 Male lung cancer death rate






20 Female lung cancer death rate


Age-Adjusted Lung Cancer Death Rates*

Per Capita Cigarette Consumption

Tobacco Use in the US, 1900–2004

10 0

1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000


Year *Age-adjusted to 2000 US standard population. Source: Death rates: US Mortality Data, 1960–2004, US Mortality Volumes, 1930–1959, National Center for Health Statistics, Centers for Disease Control and Prevention, 2006. Cigarette consumption: US Department of Agriculture, 1900–2004.

Figure 10.4. An increase in the mortality due to lung cancer in the USA followed an increase in cigarette consumption that began in the 1930s. Reprinted by the permission of the American Cancer Society, Inc. from All rights reserved.

of Americans follow this recommendation; only 24.3% eat five or more servings of fruit and vegetables daily, a figure that has changed little over the past decade [14].

The pathophysiology of cancer Now that we have examined the global burden of cancer, let’s turn to how cancers develop. While the growth and differentiation of normal cells is carefully coordinated by growth signals, cancers are characterized by uncontrolled growth and spread of abnormal cells. Unlike normal cells, cancer cells continue to grow in the absence of external growth signals, and they ignore signals to stop dividing, to specialize or to die [15]. Cancer cells can not only sustain themselves, they can expand and migrate. Normal cells can be transformed into cancer cells by a number of factors, including exposure to external carcinogens such as tobacco smoke, certain chemicals, or ionizing radiation, exposure to certain infectious agents, and exposure to certain hormones. At the cellular level, the process of cancer development is remarkably similar in many different tissues. More than 85% of cancers arise in the epithelial tissues that line our organs, such as the skin, the digestive

Figure 10.5. Cartoon of squamous epithelium. Multiple layers of epithelial cells sit atop a basement membrane. Cells become progressively more specialized as you move from the basement membrane toward the epithelial surface.

tract, the respiratory tract, and the genitourinary tract. Figure 10.5 shows a cartoon of one type of epithelial lining. In many tissues, the epithelial surface consists of multiple layers of epithelial cells. These cells sit on top of a special membrane called the basement membrane. Beneath the basement membrane there are layers of muscle and connective tissue that give the organ its structural stability. The epithelium is exposed to

Technologies for early detection and prevention of cancer


Part I: Peace Corps and rice visits: July 2–July 4, 2007 Tessa Swaziland I got up around the same time as I usually do for clinic (6:45-ish) to shower and finish packing for my long awaited visit to a Peace Corps volunteer’s site. Carrie had suggested I stay with a volunteer as soon as I got here, but once WFP stuff started, I didn’t have any time to escape the clinic. We both thought it would be a good opportunity to see where the COE’s patients come from – not just physically, but culturally, emotionally, etc . . . Like, what kind of customs and beliefs exist in their communities? How are decisions made? What is the family environment like? How are orphans and abandoned children cared for? What is the system for governing? And also, this would give me an opportunity to talk with someone whose work and goals were similar to mine but who had been here much longer and worked on many more projects. Carrie gave me a volunteer’s number a couple weeks ago, and I had called her and set up a time to come. Tandi (that’s the volunteer’s Seswati name) and I met downtown by an Internet caf´e. She was very nice and happy to answer all of my questions. In what ended up being a very rushed morning (there had been some confusion with the new WFP system, and Dave called me to the clinic mid-shower) I hadn’t managed to squeeze in breakfast, so we sat down for omelets and coffee before heading out to her homestead. I was certainly glad for her company. I’d been on combies before (the vans used for public transport here), but only for short trips between the clinic and Mbabane. Finding my way from Mbabane to Manzini, switching to another combie, riding from Manzini to Siphoneni, transferring to yet another combie, and riding out to her village (name was hard to pronounce, and now I’ve forgotten it completely) would’ve been quite an adventure (possibly an unsuccessful one). When we got off the last combie, we walked by a row of gogos (old women selling fruit or other items on the side of the road) and she greeted all of them with “Sanibonani” (the “hello” you use to address a group of people). A chorus of “Yebo”s echoed back, and the exchange continued for a minute. Everyone we passed, Tandi greeted and waved to. In that particular community, they used a two hand wave, which proved difficult since we were both carrying quite a bit of stuff. Discovering I had no Seswati name, Tandi enlisted the help of the gogos in naming me. I am now officially Zandile Dlamini. “Zandile” means “too many girls,” and Dlamini is the most common last name in Swaziland. I’d say about 40% of our COE patients are Dlamini’s. We hiked for about 20 minutes to reach the homestead where she had been living for almost a year. The paths were dirt trails randomly winding and crisscrossing through brush and occasionally along pastures and homesteads. It reminded me quite a bit of the landscape and random layout of the community I lived in five years ago in Nicaragua. At her home, I met her gogo (literally translate as “grandmother”) and the children who lived there. Her babe (father of the house) wasn’t home. In fact, he rarely is there, she said. Two of his three wives reside at that particular homestead, but neither was there at the time. One was shucking corn on the other side of the nearby mountain, and I’m not sure where the other one was. The one shucking was Tandi’s “mage” (mother . . . pronounced ma-ge with a soft g sound), and I got to meet her later. She was very well educated and easy to talk to. In fact, she had met her husband when they were both studying at a university in the UK. At that time he already had two wives back in Swaziland. She had hosted Peace Corps volunteers several times before and seemed like a very good host mom. She kept trying to make Tandi stop translating. (She thought I was another PC volunteer and thought I should know the language by now.) She said all the children there were her own, but Tandi told me later that in fact, they were all children of her husband, but none actually belonged to her. They were the children of all of his girlfriends. (Keep in mind; this is a completely normal arrangement for a family. Other than the fact that they were wealthy for a rural Swazi family, this was very representative of many of the homesteads all over Swaziland. A homestead is basically a collection of homes that belong to one extended family – a gogo, a babe, several mages, and many children).


Biomedical Engineering for Global Health

Anyway, after I met the gogo (and before I met the mage) Tandi took me over to the hospital and VCT (voluntary counseling and testing facility) to see what they were like. It looked much like the Vuvulane clinic except bigger and cleaner (still nothing close to the COE standards). We talked to the VCT employees for a bit. They basically serve as a site where community members can come in and get tested for HIV. If they test positive, there are support groups and counseling available. Also, if they’ve been diagnosed and prescribed ART, they can pick up the meds there. One nice thing about this particular VCT was the fact that there was a woman who worked there whose sole responsibility was to deal with adherence. She was Swazi, and I believe she actually grew up there or nearby. She had obtained a grant for her project and was now trying to improve adherence in the community. Unfortunately, she had left for the day, so I didn’t get to speak with her. On the way back to the homestead, Tandi told me more about her experiences living in Swaziland. She said adjusting to the culture wasn’t too difficult. She didn’t really get homesick, and the community welcomed her. Apparently the last volunteer was kind of angry. Whenever Tandi had events, people would always come up to her afterwards and say, “Thank you for not yelling at me, Sisi.” (Sisi means sister and is the word everyone uses to address a young woman who isn’t married). Evidently, the last volunteer had been a yeller. The one thing Tandi said was most difficult to adjust to was the number of deaths. Every week, there are about three vigils to mourn the deceased. They last all night and end at about 7 a.m. with a funeral.

the external world; it serves as an important protective barrier, and is constantly regenerating itself. The cells

transversely and stained with hematoxylin and eosin dyes; the hematoxylin colors the nucleus purple. The

adjacent to the basement membrane are responsible for this regeneration, therefore they are the most metabolically active cells in the epithelium. Epithelial cells

cells adjacent to the basement membrane have a large N/C ratio and this ratio becomes progressively smaller as we move up through the normal epithelium.

become more specialized and mature as we move closer to the surface of the epithelium. Those cells at the top of the epithelium are dead, and will eventually be sloughed off.

An epithelial cancer begins with transformation of a single epithelial cell. As this transformed cell grows, it fails to differentiate, continuing to actively divide. When the lower 1/3 of the epithelium is filled with transformed

Thus, there is a gradient of cell differentiation throughout the epithelium with the most differentiated cells at the epithelial surface and the least differentiated cells at the basement membrane, and the morphology of normal cells differs along this gradient. As we move from the bottom to the surface of the normal epithelium, the nucleus of each cell occupies progressively less and less of the cell volume. This ratio is called the nuclear to

cells, the condition is known as low grade pre-cancer. When the lower 2/3 of the epithelium is occupied by transformed cells, the condition is known as high grade pre-cancer. Figure 10.7 shows a photograph of an oral mucosa biopsy with a high grade pre-cancer on the right. Note the cells with increased N/C ratio in the lower 2/3 of the epithelium. These lesions are called pre-cancerous lesions because they are not yet cancers, but they have

cytoplasmic ratio (N/C ratio). A large N/C ratio is characteristic of a rapidly dividing immature cell, while a small N/C ratio is characteristic of a mature, terminally differentiated cell. Figure 10.6 shows a photograph of

the potential to develop into a cancer. As shown on the left of Figure 10.7, in some cases transformed cells can break through the basement membrane, entering the stroma beneath. In this case, we no longer have an

a biopsy from the oral mucosa that has been sectioned

organized epithelium and stroma; instead the tissue is

Technologies for early detection and prevention of cancer


Figure 10.7. Photo of stained biopsy of the oral mucosa. On the right side of the biopsy, a high grade pre-cancerous lesion is present. Note the cells with large N/C ratio that occupy most of the epithelium. The left side of the biopsy shows an invasive cancer of the oral mucosa. Nests of cancer cells are intermixed with stromal tissue.

Thus in cancer, oncogenic mutations disrupt the cell cycle and the careful coordination of growth, differen-

Figure 10.6. Photo of stained biopsy of the normal oral mucosa. Multiple layers of epithelial cells sit atop a basement membrane. Cells become progressively more specialized as you move from the basement membrane toward the epithelial surface.

a mix of nests of cancer cells and surrounding stroma. This is a significant phenomenon – it is at this point that we go from pre-cancer to invasive cancer, and as we will later see, the prognosis and treatment of pre-cancer and cancer are radically different. How do epithelial cells become transformed to initiate the development of a cancer? A cell is transformed through a series of mutations that affect its DNA; DNA mutation leads to production of mutant proteins. For example, a mutated gene can produce a defective protein that causes the growth factor receptors on the cell’s surface to be constantly on. This type of ‘gain of function’ mutation can produce a transformed cell which continues to grow in the absence of external signals. The DNA in cancer cells can also undergo mutations which result in a loss of function; for example, ignoring signals to stop growth, or losing the function to repair or destroy defective cells.

tiation and death that characterizes normal cells and tissues. Cancer cells don’t respond to signals that regulate cell growth and division. They can grow in the absence of signals to grow, and they ignore signals to stop growth. Changes in the gene expression profile allow cancer cells to replicate indefinitely. For example, normal cells can divide a finite number of times. Cancer cells overcome this limitation to become immortalized. Normal cell division is regulated by telomeric DNA at chromosome ends (Figure 10.8). This DNA functions to prevent end to end fusion of chromosomes. Normal cells shorten telomeric DNA with each division. After a certain point telomeres fuse and cells die. Most cancer cells activate an enzyme called telomerase. This enzyme extends the telomeres so that the cell can go through unlimited cycles of cell division [16]. The mutations that lead to cancer start in one cell, and as this cell divides, further mutations can occur in daughter cells. It is the accumulation of mutations that irreversibly transforms a normal cell into a cancer cell. Usually five to seven mutations are required to transform a cell. These mutations accumulate over time, which is why cancer is more common with increasing age. Fewer than 10% of mutations that lead to cancer are inherited; most are due to environmental factors [16].


Biomedical Engineering for Global Health

Learn more about it A comprehensive overview of cancer biology can be found at The section “Pathways to Cancer” contains 3D animation that illustrate the abnormalities in signaling pathways associated with cancer cells. Courtesy of Dolan DNA Learning Center, Cold Spring Harbor Laboratory.

We can think of tumor development as analogous to Darwinian evolution. The transformation of a normal cell into a cancer cell occurs via a succession of genetic changes; together, these changes lead to the progressive conversion of normal cells to cancer cells [15]. Because

ous and leaky; we will see later that we can exploit these properties to aid in the early detection and treatment of tumors. When cancer cells are confined to the organ in which they originated, we refer to the lesion as a primary

these changes confer a growth advantage relative to normal tissue, the unchecked growth of cancer cells results in a mass of tumor cells. Once a nest of transformed cells begins to grow, the energy demands of these cells

tumor. Cancer cells have the ability to spread beyond the primary organ site (Figure 10.10). As cancer cells invade through the connective tissue in the primary organ site, they can intravasate into blood vessels and

rapidly outstrip the capacity of the normal vasculature to supply nutrients [16]. As a result, transformed cells can induce the formation of new blood vessels, in a process called angiogenesis (Figure 10.9). Angiogenesis can occur in the early pre-cancerous stages. Frequently, the vessels formed in a tumor are abnormal – they are tortu-

lymph vessels in that organ. From there, they can travel to distant organs and extravasate out of the blood vessels to form metastatic nests of tumor cells in distant organs [17]. Some of these nests of tumor cells will survive, grow and expand to form metastatic tumors. Figure 10.10 illustrates how a single transformed cell can

Technologies for early detection and prevention of cancer


Figure 10.9. During angiogenesis, tumor cells induce formation of new blood vessels.

Why is early detection so important? In his 1971 State of Union address, President Nixon declared “war” on cancer and requested $100 million for

Figure 10.8. A fluorescent stain indicates telomeric DNA (lighter areas) at the tips of chromosomes. Courtesy of Peter M. Landsdrop, BC Cancer Research Centre.

lead to the development of a metastatic tumor. Metastasis is responsible for a large fraction (90%) of deaths due to cancer [16]. In summary, there are more than 100 different types of cancer, yet the process of tumor development is remarkably similar across different organ sites. The formation of tumors is a multi-step process, during which six essential alterations occur in cell physiology: (1) cells develop self-sufficiency in growth signals, (2) they become insensitive to signals of growth inhibition, (3) they evade programmed cell death, (4) develop limitless replicative potential (5) they can sustain angiogenesis, and (6) acquire the ability to invade tissue and metastasize [15]. As we will see later in this chapter, the development of new diagnostic and therapeutic techniques for cancer increasingly focuses on these six common elements of cancer cells.

cancer research. On December 23, 1971, Nixon signed the National Cancer Act into law and said, “I hope in years ahead we will look back on this action today as the most significant action taken during my Administration [18].” Today, the US government still makes a substantial investment in cancer research; the National Cancer Institute (NCI) will spend over $4.6 billion for cancer research in 2007 [19]. The mission of the NCI is to eliminate suffering and death due to cancer by 2015 [20]. How have cancer incidence and mortality rates changed as a result? In 1950, heart disease was the leading cause of death, followed by cancer, cerebrovascular disease, and infectious disease (Figure 10.11). More than 50 years later, in 2004, the age adjusted mortality due to heart disease, cerebrovascular disease, and infectious disease have all dropped by more than half; that due to cancer has decreased only slightly [21]. Table 10.7 shows the five year survival rates for patients diagnosed with different cancers during three different time periods: 1975–1977, 1984–1986 and 1996–2004. In the mid 1970s, the overall five year survival rate was 50% and this did not change appreciably over the next decade. In the late 1990s to early 2000s, the five year survival rates had risen to 66%. Prostate cancer


Biomedical Engineering for Global Health

Figure 10.10. Development of a metastatic tumor. Courtesy of

showed great improvement between the 1980s and 1990s, reflecting the introduction of a screening test

a distant location prior to diagnosis, the five year survival rates are dismally low. Thus, an important strategy

designed to detect early disease [1]. Lung cancer and pancreatic cancer survival rates have not changed substantially during this period; these cancers tend to be diagnosed at a relatively advanced stage when available therapies are not particularly effective. Why is early detection so important to reducing can-

to reduce the mortality associated with cancer is to develop improved detection technologies designed to identify cancer at the earliest possible stages, when the available therapies are more likely to result in cure.

cer mortality? Figure 10.12 compares the five year survival rates for three different cancers as a function of the stage at which they are diagnosed. The five year survival rates exceed 90% for patients whose cancer is

Strategies for early detection

diagnosed when it is still confined to the local organ site [1]. The five year survival rate drops when cancer is first detected after metastasis to a regional location. For those patients whose cancer has metastasized to

Typically, cancers do not produce symptoms until a fairly late stage. How can we identify disease in asymptomatic patients? This is the goal of a process called “cancer screening.” Screening refers to the use of simple tests in a healthy population. The goal of screening is to identify individuals who have disease, but do not yet have symptoms. The goal of screening is not to

Technologies for early detection and prevention of cancer


Table 10.7. Changes over time in five year survival rates [1].

American Cancer Society. Cancer Facts and Figures 2009. Atlanta: American Cancer Society, Inc.

Figure 10.11. Changes in US death rates since 1950. Source: CDC.


Biomedical Engineering for Global Health examination and screening mammography, cervical cancer with HPV testing and the Pap smear, prostate cancer using the serum PSA test and digital rectal examination, and colon and rectal cancer, using a combination of the fecal occult blood test, flexible sigmoidoscopy, and colonoscopy [24]. Table 10.8 summarizes the recommendations of the American Cancer Society regarding screening for breast cancer; yearly mammograms to screen for breast cancer are recommended for women over the age of 40. While 69% of women over age 40 report having received a

Figure 10.12. Five year relative survival rates due to several different cancers [3].

diagnose disease, but rather to use an inexpensive and simple test to identify those individuals who should have a more expensive and accurate test to confirm a diagnosis of disease. In the USA, we routinely screen for four cancers, including female breast cancer with clinical breast

mammogram, women with no health insurance are significantly less likely to have received a mammogram. Table 10.9 summarizes the recommendations of the American Cancer Society regarding screening for cancers of the colon and rectum. While the percentage of people age 50 or more reporting a recent flexible sigmoidoscopy has increased in the past few years, only 45% of patients have had this test. The percentage of patients with no health insurance reporting this screening test is only 17% [14].

Table 10.8. Recommendations for breast cancer screening from the American Cancer Society. Reprinted with permission from the American Cancer Society, Inc. from All rights reserved.

Table 10.9. Recommendations for screening for cancers of the colon and rectum from the American Cancer Society. Reprinted with permission from the American Cancer Society, Inc., from All rights reserved.

Technologies for early detection and prevention of cancer


Cancer pain management: China Cancer pain: it is a component of the disease that many patients fear more than death itself. Its severity has been described as intolerable and excruciating, and it only increases with the progression of the cancer. While cancer pain is treatable with the use of standard analgesics, including opioids such as morphine, this cheap and effective analgesic is not the standard of care everywhere. “It was clear that many places didn’t even have aspirin for cancer pain relief. And in many countries, the idea of having more potent drugs, even codeine let alone morphine, which is what we use to manage severe cancer pain, was nonexistent,” recalls Dr. Charles Cleeland, Chair of the Department of Symptom Research at the University of Texas M.D. Anderson Cancer Center (image courtesy of Dr. Cleeland). In 1993 the morphine consumption in China was estimated at 0.01 mg per capita, a surprisingly low number compared to the estimated 66.53 mg per capita consumption for Denmark [22]. Morphine was strictly managed. Suffering Chinese patients needed a certificate to receive morphine, and could only receive one ampoule of short-lasting morphine per day for four days as long as they returned an empty ampoule every time [22]. Patients had to choose the hour of their pain relief. Such telling statistics urged the World Health Organization (WHO) to initiate a plan to improve cancer pain management globally. Up to this point, Dr. Cleeland, whose work focuses on pain assessment and treatment, had worked in countries such as Mexico and the Dominican Republic, instructing small groups in the usage of opioids for cancer pain relief. “At that time I had a postdoctoral fellow from China who said, ‘Why don’t you take on something really big?’ So she got me connected with a friend of hers who was the minister of the health for the district of Beijing. So he came over and we tried to think of things to do to start.” Thus began a collaboration between the WHO, the government of China and Dr. Cleeland’s pain research group to address the inadequate cancer pain management in China. The initial step was to study the epidemiology of the problem. To do this Dr. Cleeland launched a study led by Dr. Shelley Wang in 1992. The study helped elucidate the current state of pain control in the greater area of Beijing by examining 200 different cases. “And what we found was not a surprise; morphine was to be used rarely if at all. These patients had very high levels of pain compared to more developed countries.” To try and explain the kind of pain the group encountered, Dr. Cleeland says, “On a 0 to 10 scale, if we break that scale up you and I probably experience from time to time pain maybe up to 3 or 4. Many cancer patients after their disease has metastasized have a continual 10 pain.” The lack of pain control was evident, and the reasons behind this were many. “There were regulatory problems. There were issues of concern with addiction and a lack of any kind of distribution of controlled substances. It was a multifaceted problem.” The next step was to set up large meetings in Beijing and other major cities to introduce the epidemiology of the problem and to begin instruction. Over 500 individuals selected by government officials attended these three-day meetings. “They chose very well. They picked people who were administrators for major hospitals. They picked drug regulators. They picked nursing people as well as physicians.” Once the problem was discussed, a series of “trainer training” programs began. Groups composed of a nurse, a pharmacist and a physician from different hospitals would attend the program and learn about pharmacology of opioids and pain assessment. “The impressive thing, always, was when we’d train the students and then


Biomedical Engineering for Global Health

send them into the wards. They would ask the physicians how many patients have pain and the physicians would say “Well, of course, none of my patients have pain.” Then they would go around and ask and find it was quite a different story.” Gradually, the training was handed over to Chinese professionals. To support their work, an evidence-based text developed by the WHO in the subject of cancer pain treatment was translated to Chinese. Concomitantly, the government of China adopted a new cancer pain relief policy, adjusting the inhibiting narcotics control policy, approving new opioid analgesics for sale and distribution, and increasing opioid manufacturing through joint ventures and other means. Gradually the pain alleviation for suffering cancer patients spread throughout the country. Five years later, a comparison study was done. “It was a tremendous change from the majority of patients being under-managed to a majority of patients being managed very well according to the WHO standards. It was a 50% drop, from 60% under-managed in 1992 to 30% under-managed in 1997,” Dr. Shelley Wang recalls. Today opioids such as immediate and sustained-release morphine, fentanyl patch, meperidine, methadone and codeine are available to cancer patients in China. Physicians can prescribe these analgesics for up to 5 days, and in the case of more severe pain a physician can prescribe 15-day relief in the form of the fentanyl patch [23]. And, although the work is still not complete, a network of committed policy makers, oncologists and pain experts have been able to tremendously change the excruciating pain experience for a nation. Unfortunately, the beginning of this story is not uncommon. “You have about 10 million cancer deaths a year worldwide, and probably 2/3 of them will experience significant pain. And, unfortunately, at least more than half will be inadequately managed. So they’ll have a period of 3 or 4 months at the end of life which will be just miserable,” states Dr. Cleeland. Presently, Dr. Cleeland, Dr. Wang and their group continue to address the cross-cultural issue of pain and agree that its poor treatment is due to the lack of appreciation and assessment of pain. They hope to reproduce the successful model established in China and are involved in joint efforts to help other nations such as Russia, Korea and Japan, bring relief to cancer patients [22, 23].

Effectiveness of screening How do we judge the effectiveness of a screening test? Let’s take the example of screening for breast cancer. Imagine that you are a patient being screened with mammography. We can envision four possible results. If you have breast cancer and the mammogram is positive, the result is a “true positive.” However, if you have breast cancer but the mammogram is negative, the result is a “false negative.” If you do not have breast cancer, and the test is negative, the result is a “true negative.” Finally, if you do not have breast cancer, but the test is positive, the result is a “false positive.” We can arrange these possible outcomes in a 2 × 2 table as shown in Figure 10.13. We define the sensitivity of a test as the probability that given DISEASE, the patient tests POSITIVE. The sensitivity is a measure of the ability of the test to cor-

rectly detect disease when it is present, or the ability to find true positives. Sensitivity can range from a low of 0% to a high of 100%. We define the specificity of a test as the probability that given NO DISEASE, the patient

Mammogram Positive

Mammogram Negative

Patient Has Breast cancer

True Positive (TP)

False Negative (FN)

Patient Does Not Have Breast Cancer

False Positive (FP)

True Negative (TN)

Figure 10.13. Possible outcomes of a screening mammogram for breast cancer.

Technologies for early detection and prevention of cancer Table 10.10. Comparing the effectiveness of a breast exam versus a mammogram in screening for breast cancer [25].


Test Positive

Test Negative

Disease Present

TP = 11

FN = 2

Number with Disease = TP + FN = 13

Disease Absent

FP= 15

TN = 208

Number without Disease = FP + TN = 223

Number who Test Positive = TP + FP = 26

Number who Test Negative = FN + TN = 210

Total Number Tested = TP + FN + FP + TN = 236

Figure 10.15. Data to calculate the sensitivity and specific of MRI screening for breast cancer. Used with permission from [26].

The specificity can be calculated as: Sp = Test Positive

Test Negative

Disease Present



Number with Disease = TP + FN

Disease Absent



Number without Disease = FP + TN

Number who Test Positive = TP + FP

Number who Test Negative = FN + TN

Total Number Tested = TP + FN + FP + TN

Figure 10.14. Possible outcomes of a diagnostic or screening test.

tests NEGATIVE. Specificity characterizes the ability of a test to avoid calling normal things disease, or the ability to avoid false positives. Specificity can also range from a low of 0% to a high of 100%. A perfect test has a sensitivity of 100% and a specificity of 100%. If a test performs better than chance alone (or better than the toss of a coin), the sum of the sensitivity and specificity is greater than 100%. Table 10.10 lists the average reported sensitivity and specificity of two different screening tests for breast cancer – clinical breast exam, and mammography. The average sensitivity of mammography is 75%. This is higher than the 54% sensitivity of clinical breast exam. The specificity of mammography is 92%, slightly lower than that of clinical breast exam [25]. How do we measure the sensitivity and specificity of a screening test? If we screen a population of patients, some of whom are known to have disease and some who are known to be disease free, we can calculate the sensitivity and specificity of the test. Figure 10.14 shows the possible outcomes of the testing. The sensitivity can be calculated as: Se =

TP TP = . (number with disease) (TP + FN)


TN TN = . (number without disease) (TN + FP)


As an example, let’s calculate the sensitivity and specificity of a new test suggested for breast cancer screening – magnetic resonance imaging or MRI. In 2004, results from a clinical trial of 236 women were reported to assess the performance of MRI for screening for breast cancer in women at high risk of developing the disease [26]. During the first year of the study, each woman had an MRI exam; 26 women had an abnormal MRI. To confirm the presence of breast cancer, additional testing was performed in women with an abnormal breast MRI; these tests confirmed the presence of breast cancer in 11 women. An additional two women who had a normal MRI exam were found to have breast cancer based on other tests. Figure 10.15 shows the 2 × 2 table filled out for this example. We can calculate the sensitivity and specificity of MRI in this clinical trial as follows. Se =

(11) TP = = 84.6% (TP + FN) (11 + 2)


Sp =

(208) TN = = 93.3% (TN + FP) (208 + 15)


In order to calculate the sensitivity and specificity of a new test, we must develop criteria to determine whether the test result is normal or abnormal. As these criteria change, our estimate of the test’s sensitivity and specificity change. We can characterize the performance of a test by plotting the test sensitivity and specificity as we vary these criteria. The resulting plot of sensitivity vs. specificity is known as a receiver-operator characteristic curve (ROC curve). Figure 10.16 shows the ROC curve for MRI used to screen for breast cancer in high risk women calculated from the study above [26]. As the


Biomedical Engineering for Global Health Table 10.11. MRI is significantly more expensive than other clinical methods of breast cancer screening [25].

Screening Method Clinical Breast Exam Mammography Ultrasound MRI

Medicare Reimbursement $39 $90 $74 $1108

As a patient, you wish to be screened with a test that has both a high sensitivity and specificity. But how high do these values need to be for the test to be useful to you? If you receive a positive screening test result, what is the likelihood that the result is a true positive or a false positive? Similarly, if you receive a negative screening test result what is the likelihood that the result is a true negative or a false negative? The sensitivity and specificity of the test don’t provide enough information to answer these questions. Instead, we must calculate the positive and negative predictive value of the test, which give these probabilities. The positive predictive value (PPV) is the probability that, given a POSITIVE test result, you have DISEASE. PPV ranges from 0% to 100%. The negative predictive value (NPV) is the probability that given a NEGATIVE test result, you do NOT HAVE DISEASE. Again, NPV Figure 10.16. ROC curves for different breast cancer screening modalities. The area under the curve is highest for MRI. Used with c permission from [26]. JAMA, 2004, 292; 1317–25. Copyright  2004 American Medical Association. All rights reserved.

sensitivity increases, the specificity decreases. The area under the ROC curve is often used to provide a measure of test accuracy. A perfect test has an ROC curve with area 1; a test that performs no better than chance has an area under the curve of 0.5. Figure 10.16 also compares the ROC curves for several screening methodologies in the same group of patients; the area under the ROC curve is highest for MRI. Unfortunately, the cost of MRI is substantially higher than the cost of clinically accepted technologies (Table 10.11) [25]. In Chapter 11, we will examine how to decide whether the additional resources required to implement a new technology represent a good investment.

ranges from 0% to 100%. We can use our 2 × 2 table to calculate PPV and NPV. TP TP = (# testing positive) (TP + FP) TN TN NPV = = (# testing negative) (FN + TN) PPV =

(10.5) (10.6)

Again, we can use our MRI example of Table 10.11 to illustrate how to calculate positive and negative predictive value. PPV =

(11) = 42% (26)



(208) = 99% (210)


These statistics tell a patient that, given an abnormal screening MRI, there is only a 42% chance that she actually has breast cancer. Further testing is required to confirm whether cancer is actually present. However,

Technologies for early detection and prevention of cancer


Breast cancer screening in China: is breast self-examination an alternative to mammography? While screening mammography can reduce the mortality associated with breast cancer, it is not available in many developing countries due to limited resources. In such settings, breast self-examination may provide a less expensive alternative. A recent randomized clinical trial of breast self-examination was conducted in Shanghai, China to determine whether breast self-examination could reduce breast cancer mortality. Beginning in 1989, more than 266,000 women in Shanghai were randomized into two groups. One group of 132,979 women received initial instruction in breast self-examination, with reinforcement sessions one and three years later. These women practiced breast self-examination every six months for five years; and 133,085 women were assigned to a control group. All women were followed through December 2000 for mortality from breast cancer. The graph shows the cumulative breast cancer mortality per 100,000 women in the breast self-examination group (solid line) and the control group (dashed line). There were a total of 135 breast cancer deaths in the group who received instruction in breast self-examination, compared to 131 breast cancer deaths in the control group. In addition, more benign breast lesions were discovered in the group of women who performed breast selfexamination. Unfortunately based on this study, it does not appear that breast self-examination can reduce mortality from breast cancer in this setting [27].

From [27] with permission from Oxford University Press.

given a normal screening MRI, there is a 99% chance that the patient truly does not have breast cancer. Clearly, the NPV and the PPV of a test depend on the sensitivity and specificity of the test. But they also depend on the prevalence of the disease that we are screening for. Prevalence is a measure of whether a disease is common or rare. Recall that prevalence of disease in a population, p, is defined as: (# in population with disease) . p= (total # in population)


(p)(Se) [(p)(Se) + (1 − p)(1 − Sp)]



(1 − p)(Sp) [(1 − p)(Sp) + (p)(1 − Se)]


Using our example of MRI to screen for breast cancer again, the prevalence of disease is: p=


In terms of our 2 × 2 table, prevalence can be calculated as: (TP + FN) . p= (TP + FP + TN + FN)


(13) = 0.055 or 5.5%. (236)

We can use the formulas above to calculate positive and negative predictive value. PPV =

(0.055)(0.846) = 42% [(0.055)(0.846) + (0.945)(0.067)] (10.14)


(0.945)(0.933) = 99% [(0.945)(0.933) + (0.055)(0.154)] (10.15)


If we know the prevalence of a disease and the sensitivity and specificity of a test, we can calculate the positive and negative predictive values of the test as



Biomedical Engineering for Global Health

We obtain exactly the same results as calculated previously. As the prevalence of disease decreases, the PPV of a test decreases. In our example of MRI to screen for breast cancer, the study was designed to screen women who were at high risk for breast cancer. As a result, 5.5% of the population studied had breast cancer, a prevalence that is much higher than that in the general population. Let’s consider what would happen to the predictive value of the test if we were to use the same test to screen for breast cancer in all women. In the United States approximately 131 new cases of breast cancer are identified per 100,000 women [28]. Let’s calculate the prevalence, positive and negative predictive value under these conditions. The prevalence is: p=

(131) = 0.0013 = 0.13%. (100, 000)


The sensitivity and specificity are independent of disease prevalence, and as before are:

cancer. Among women between the ages of 40–49 years, we must screen 1792 women for 14 years to prevent one death from breast cancer [29]. With this introduction, we now examine three cancers in detail – cervical cancer, prostate cancer, and ovarian cancer. In each case, we will examine the efficacy of existing screening technologies. We will also examine the new technologies in development to improve early detection.

Early detection of cervical cancer In 2007, there were predicted to be 11,150 new cases of cervical cancer in the USA, and 3670 deaths due to cervical cancer [1]. Worldwide, cervical cancer is an important problem. In 2002, 493,000 new cases of cervical cancer were reported globally. Some 83% of cervical cancers occur in the developing world, with the highest incidence in central and South America, southern Africa and Asia (Figure 10.17). Cervical cancer caused 274,000 deaths

Se = 84.6%


Sp = 93.3%.


in 2002 worldwide, and was the leading cause of female cancer mortality in developing countries [6]. Cervical

However, the positive and negative predictive values

cancer affects relatively young women, and is the single largest cause of years lost to life due to cancer in the developing world [30]. The cervix is located between the vagina and the uterus (Figure 10.18). The cervical os is the opening

differ. PPV =

(0.0013)(0.846) = 1.6% [(0.0013)(0.846) + (0.99987)(0.67)] (10.19)

into the uterus; during conception, sperm travel from (0.99987)(0.933) = 99.98% NPV = [(0.99987)(0.933) + (0.0013)(0.154)] (10.20) The PPV deceases substantially, while the NPV increases slightly. Suppose a woman in our study has a positive MRI. What is the likelihood that she has breast cancer? This is the same as the positive predictive value and is only 1.6%! The low PPV illustrates the challenge of screening for a rare disease. For this reason, we generally screen for breast cancer in older women, because the prevalence of breast cancer increases with age. While screening for breast cancer using mammography clearly reduces breast cancer mortality, it has been estimated that we must screen 1224 women for more than 14 years in order to prevent one death from breast

the vagina through the os to fertilize an egg. Throughout a pregnancy, the cervix provides the structural stability to hold the fetus inside the womb. During labor and delivery, the cervix thins and stretches to enable the baby to travel through the birth canal. Thus, the wall of the cervix contains both collagen and elastin fibers to provide both strength and elasticity. The outer surface of the cervix comes into contact with both semen and potentially dangerous bacteria and viruses. The cervix is lined with multiple layers of epithelial cells that play an important role both in preventing infection and facilitating conception. These epithelial cells produce cervical mucus; the mucus changes consistency throughout the menstrual cycle in order to facilitate travel of sperm during ovulation and to prevent travel and growth of pathogens.

Technologies for early detection and prevention of cancer


Figure 10.17. Global predictions for cervical cancer incidence in 2005.

squamo-columnar junction (SQ junction). Most cervical cancers begin when an epithelial cell in the squamous epithelium becomes transformed and begins to proliferate. When an epithelial cell near the basement membrane becomes transformed, it loses the capacity to terminally differentiate, and the epithelium gradually fills with actively dividing cells that have large nuclei. When the bottom 1/3 of the epithelium is filled with transformed cells, the condition is referred to as a low grade squamous intraepithelial lesion (LGSIL). When the bottom 2/3 of the epithelium is filled with transformed cells, the condition is referred to as a high grade squaFigure 10.18. Anatomy of the female reproductive tract.

Cervical cancers begin in the epithelial lining of the cervix. Two types of epithelial tissue line the cervix (Figure 10.19); in both cases the epithelial cells are separated from the supporting stromal tissue below by a thin basement membrane. Surrounding the os, the surface of the cervix has small finger-like projections and is lined by a single layer of columnar epithelial cells. The outer edges of the cervix are flat and lined by multiple layers of squamous epithelial cells. The squamous epithelium is typically 200–300 microns thick. The junction between the columnar and squamous epithelium is known as the

mous intraepithelial lesion (HGSIL), and when the complete epithelium is transformed, it is called a carcinoma in situ (CIS). At this stage, the lesions are considered to be pre-cancerous. However, if the transformed cells break through the basement membrane and migrate into the stroma, the condition is known as a micro-invasive cancer. This cycle is known as the pre-cancer to cancer sequence. The prognosis of micro-invasive cancer is much more serious and the treatment is much more invasive than that of pre-cancers. Thus, the focus of cervical cancer screening programs is to identify cervical pre-cancers (when they can be easily treated) before they become cervical cancers (when they are difficult, painful and expensive to treat). Most low grade precancers regress on their own, while 20–45% of high


Biomedical Engineering for Global Health

Figure 10.19. A cross section (left) showing the two types of tissues lining the cervix; in normal cervical tissue (right) squamous cells change in appearance from top to bottom (A), but in pre-cancer this gradient becomes increasingly absent. Reprinted from Joumal of Nurse-Midwifery, 3b(5), McGraw R.K., c 1991, Gynecology: A Clinical Atlas  with permission from Elsevier.

grade lesions progress to cervical cancer if untreated. The progression from pre-cancer to cancer has been esti-

In a wart or benign infection, the HPV chromosomes are

mated to take about a decade [31].

stably maintained in the basal epithelium as plasmids whose replication keeps step with the chromosomes of the host (Figure 10.21, left). A cell becomes trans-

What causes transformation of cervical epithelial cells?

formed when the viral DNA is integrated into a host chromosome. This alteration of the viral gene environment can disrupt control of their expression. The unreg-

causative factor in squamous cell carcinoma of the cervix. HPV infection is the most common sexually transmitted disease; asymptomatic HPV infections can be detected in 5–40% of women of reproductive age [32]. The majority of women with HPV infection do not develop invasive cervical cancer. In most young women, HPV infections are transient; the immune system clears them with no ill effects. However, if HPV infection persists past age 30, there is a much greater risk of developing cervical cancer [33]. There are more than 100 different types of the human papillomavirus; not all of them are carcinogenic. Fifteen types of HPV are commonly linked to cervical cancer, with HPV types 16 and 18 the most commonly found high risk types of virus (Figure 10.20) [34]. Human papillomaviruses have double stranded circular DNA chromosomes with about 8000 nucleotide pairs [35]. In an HPV infection, the HPV genetic material is transported to the nucleus of infected cervical epithelial cells.

ulated production of viral proteins tends to increase the Global Prevalence of HPV Types in Cervical Cancer 16 +18 +15 +31 HPV Types

In the 1990s, researchers demonstrated that infection with human papillomavirus (HPV) is the central


57.6% 71.7% 77.4% 81.3% 85.0%













Figure 10.20. Fraction of cervical cancer accounted for by different types of HPV. The two vaccines under development and testing protect against HPV types 16 and 18, which together account for about 70% of cervical cancers [34]. Credit: X. Bosch And N. Munoz/Iarc, Ibsccs, And Multicentric Studies (N = 3045). From Science 29 April 2005: Vol. 308. no. 5722, pp. 618–621. Reprinted with permission from AAAS.

Technologies for early detection and prevention of cancer


Figure 10.21. When HPV DNA is integrated into the host cell DNA, growth is no longer regulated and a malignant tumor can form.

rate of cell division, thereby helping to generate a cancer (Figure 10.21, right). For example, the HPV E6 protein appears to alter cell growth through effects on p53, an endogenous tumor-suppressor protein. E6 binds to p53, targeting it for destruction [35]. The signs and symptoms of cervical cancer include abnormal vaginal bleeding, in between periods or especially related to intercourse, and pelvic pain. Advanced cervical cancer is treated using surgery, and a combination of radiation therapy and chemotherapy. In the USA, the five year survival rate for localized cervical cancer is excellent at 92%. Slightly more than half of cervical cancers in the USA are diagnosed at this stage [1]. Such a large fraction of cervical cancers are detected early because we have a good screening test. In fact, most lesions are caught at the stage where they are still pre-cancers, and can be treated easily before they progress to cancer.

How do we detect cervical cancer and its precursors? We screen for cervical cancer and its precursors using a test called the Papanicoloau (Pap) smear. The use of the Pap smear to screen has resulted in dramatic decreases in the incidence and mortality of cervical cancer, and is largely viewed as the most successful cancer screening test in medical history. The diagnosis of cervical cancer and its precursors is made using a confirmatory followup test, colposcopy and biopsy.

Figure 10.22. A wooden spatula is used to obtain a Pap smear sample. By permission of the Mayo Foundation for Medical Education & Research. All rights reserved.

In a Papanicoloau smear, a speculum is inserted into the vagina enabling the healthcare provider to visualize the cervix. A small wooden spatula is scraped against the cervix; the spatula is placed at the squamocolumnar junction and rotated to scrape off epithelial cells all around the junction (Figure 10.22). The cells collected on the spatula are then smeared on to a glass slide and allowed to dry. In obtaining a successful Pap smear, more than 50,000–300,000 cells, including both columnar and squamous epithelial cells, will be placed on the slide. The cells are then stained and examined by a trained cyto-technologist. Any abnormal appearing cells (cells with large nuclei or abnormal chromatin)


Biomedical Engineering for Global Health

US regulations to ensure quality in cytology laboratories In the mid 1980s, a number of cases were brought to light in which women had developed cervical cancer despite having routine Pap smears with Figure 10.23. Normal appearing cells (left) and abnormal appearing cells (right) in a Pap smear. Copyrighted and used with permission of Mayo Foundation for Medical Education and Research, all rights reserved.

are noted (Figure 10.23). Based on these changes, Pap smears are classified into several categories: normal, infection/repair, atypical cells of uncertain significance, low grade pre-cancer, high grade pre-cancer, and cancer. Interpretation of the Pap smear is subjective, and the reproducibility of this interpretation has been found to be poor. An individual clinician agrees with their own prior diagnosis about 78% of the time and agrees with the diagnosis of others only between 28–72% of the time

normal results. Media coverage of the cases implied that the false negatives resulted from laboratory errors due to carelessness. At the time, many commercial clinical laboratories set very high target rates for cytotechnologists to screen a certain number of slides per day or risk being fined part of their salary. Several such cases received extensive media coverage. As a result, the US Congress passed the Clinical Laboratory and Improvement Amendments of 1988 (CLIA). CLIA limited the number of slides that cytotechnologists in the USA can review to no more than 100 slides per day. CLIA also mandated that 10% of slides with a “normal” diagnosis be re-screened in order to limit the number of false negative diagnoses [37–39].

[36]. While Pap smears are helpful in identifying cervical cancer and its precursors, a number of factors can lead to false positive and false negative results. Only a small fraction of Pap smears contain abnormalities, and in those cases only a small percentage of cells may show cancerous changes, so positive cases are

of the Pap test ranges from 30% to 87% (average 47%), while the specificity ranges from 86% to 100% (average

sometimes missed due to human error. Since the Pap

results with biopsy [41]. The sensitivity and specificity vary widely from one study to another, and it is clear that it is difficult to achieve simultaneously high sensitivity and specificity using the Pap test. Because the Pap smear is a screening test, an abnormal Pap smear is usually followed by a diagnostic

smear samples only a fraction of cells from the cervix, abnormal cells present on the cervix may not be exfoliated when the sample is collected. Finally, benign changes such as infection or inflammation and tissue repair can cause cells to have the appearance of precancerous cells. It is difficult to measure the accuracy of the Pap smear for several reasons. For example, in many studies of Pap test accuracy, only patients with an abnormal Pap smear receive further testing to confirm the presence of disease. Studies of this type suffer from what is known as verification bias; only enough data are collected to allow one to calculate the specificity of the test, but not the sensitivity. Recent studies designed to verify both positive and negative results indicate that the sensitivity

95%) when low grade Pap smears and worse are considered to be abnormal [40]. Figure 10.24 shows sensitivity and specificity of 62 different studies comparing Pap test

procedure called colposcopy. In colposcopy, a speculum is inserted and a low power microscope (called a colposcope) is used to view the cervix (Figure 10.25a,b). A solution of weak acetic acid (vinegar) is applied to the cervix. The vinegar washes away any cervical mucus and also causes any pre-cancerous areas of tissue to turn white. Any suspicious areas on the cervix are then biopsied, using a metal biopsy forceps to remove a peasized portion of tissue. The biopsy is cut, stained and examined under the microscope by a pathologist. The

Technologies for early detection and prevention of cancer




Figure 10.24. The sensitivity and specificity of the Pap smear from 62 different studies can be used to estimate the ROC curve of the Pap test. From [41]. Used with permission from Oxford University Press.

sensitivity of visual examination using the colposcope is excellent, at 96%; however, the specificity is quite low, only 48% [42]. The low specificity of colposcopy is the reason that a confirmatory biopsy must be obtained; but due to the low specificity, more than half of all biopsies obtained at colposcopy show only benign changes. Figure 10.26a shows a histologic section of normal cervix from the squamous epithelium prepared from a biopsy obtained under colposcopic guidance; the cervix is lined by about 10–15 layers of epithelial cells. In the normal cervix, the basal layer of cells has the largest N/C ratio. The cells at the top of the epithelium have the smallest N/C ratio. Figure 10.26b shows a histologic section of a high grade pre-cancer. Clearly, the N/C ratio is increased throughout the entire epithelium. In this specimen, the cells have not yet invaded the basement membrane to form a micro-invasive cancer. In summary, we screen for cervical cancer and its precursors using the Pap smear, and we confirm the diagnosis using colposcopy and biopsy. Pre-cancerous cervical lesions can be removed using a simple outpatient electro-surgical procedure to remove the transformed epithelium; this treatment preserves fertility. Because

Figure 10.25. (a) The use of a colposcope to view the uterine cervix. (b) Colposcopic photo of cervix.

we have a good screening test, most lesions are caught at the stage where they are still pre-cancers, and can be treated easily before they progress to cancer. As a result, by screening for pre-cancer, we can actually reduce the incidence of cervical cancer. Before the introduction of screening programs, the incidence of cervical cancer in North America and Europe was similar to that seen in developing countries today. In every country in which organized screening programs based on the Pap smear have been introduced, rates of cervical cancer incidence and mortality have decreased [30]. While cervical cancer screening

Biomedical Engineering for Global Health




30 20



1955 1960 1965 1970 1975 1980 1985 1990 1995 YEAR Denmark




Figure 10.27. Incidence rates of cervical cancer in four Nordic countries. Decreases in the incidence rate parallel the introduction and extent of screening programs. The article was published in Vaccine, Vol. 24S3, D. Maxwell Parkin and Freddie Bray, The burden c Elsevier (2006). of HPV-related cancers, pp. S3/11–S3/25,  (b)

has never been tested in randomized clinical trials, reductions in cervical cancer mortality and incidence in countries where screening is practiced provide evidence that screening is effective. Figure 10.27 shows the declines in cervical cancer incidence in Nordic countries where screening programs were introduced in the 1960s to 1970s. Figure 10.28 compares changes in incidence rates over time in four countries. Decreases in Shanghai, China, reflect the introduction of an intensive screening program; in contrast, incidence rates have been stable in Bombay, India, where screening is largely unavailable [30]. The Pap smear is viewed as one of the most successful public health measures ever introduced. Given the relatively low sensitivity and specificity of the Pap test, it is sometimes surprising that screening has been

Figure 10.26. (a) Photograph of a normal cervical biopsy. (b) Photograph of high grade pre-cancer.

so successful in reducing the incidence and mortality of cervical cancer. In large part, this is because, on average, to go from cervical pre-cancer to invasive cervical cancer requires a decade [44]. Even if a woman has a falsely negative Pap smear one year, chances are it will be detected when she has her next Pap smear, before it has progressed to cancer. However, the Pap smear does miss some cervical cancers. It has been estimated that

Technologies for early detection and prevention of cancer

30 40 50


the USA following up low grade Pap smears that likely will not yield any health benefits [46]. Because of the costs and infrastructure requirements associated with the test, the Pap smear is not available to a large segment of the world’s population and cervical


cancer continues to kill many young women. There are many barriers to cervical cancer screening in develop-



ing countries; developing countries face a lack of trained cytotechnologists and cytology labs. A further complication is the lack of facilities to follow up abnormal Pap smears and treat pre-cancerous lesions. It is difficult for women who live in rural areas to come for multiple visits required to screen, diagnose and treat cervical cancer and its precursors. Finally, the costs of screening in many developing countries exceeds a family’s daily income, putting the test out of reach for most [47]. A number of new technologies have been developed to address the limitations of the Pap smear. In large part, these technologies have three goals: (1) reduce the false positive and false negative rates of the test, (2)




develop tests that can give instantaneous results so that women could be treated at the initial visit if the test were positive, and (3) reduce the costs of the test so that it can be implemented in the developing world. Here, we will


examine four new technologies to screen for cervical cancer: liquid based cytology, automated Pap smears, HPV testing, and optical testing. 1960




2000 YEAR

China, Shanghai Colombia, Cali India, Bombai Japan, Miyagi

Figure 10.28. Incidence rates of cervical cancer in four countries. Where screening programs are not available, incidence rates have been stable. The article was published in Vaccine, Vol. 24S3, D. Maxwell Parkin and Freddie Bray, The burden of HPV-related c Elsevier (2006). cancers, pp. S3/11–S3/25, 

3% of preventable cervical cancer deaths can be traced to false negative readings; an estimated 50% of these are due to sampling error that would not be detected with re-screening [45]. We spend more than $6B annually in

Liquid cytology One of the primary limitations of the conventional Pap smear is that only a fraction (estimated to be 20%) of the cells collected are transferred onto the slide which is later stained and examined for transformed cells [44]. A new technique called thin layer, liquid based cytology has been developed to improve the conventional Pap smear. In this technique, the brush used to collect the Pap smear is dipped into a vial containing a cell preservative, ensuring that all cells which are collected are available for analysis. A robotic preparation device is then used to remove blood and inflammatory cells and transfer a thin layer of representative cells in a circular area onto a slide. In one liquid cytology procedure R (ThinPrep ), a spinning cylinder is lowered into the specimen vial and used to break up any clumps of cells.


Biomedical Engineering for Global Health

How often should women be screened using the Pap smear? The American Cancer Society issued Screening Guidelines for the Early Detection of Cervical Cancer in 2002. These guidelines recommend that screening should begin approximately three years after women begin having vaginal intercourse, but no later than 21 years of age. Screening should be done every year with regular Pap tests or every two years using liquid based cytology. If a woman has had three normal tests in a row and has reached 30 years of age, she may reduce the frequency of screening to every two to three years. However,

R Pap and (b) conventional Pap smear. Figure 10.29. (a) ThinPrep Used with permission from [44]. Reprinted with permission from Elsevier (The Lancet Oncology, 2001, Vol. 2 No. 1, pp. 27–32).

The cell suspension is then drawn upward through a polycarbonate filter until an approximate single layer of cells covers the filter. The filter is then briefly adhered to a glass slide in order to transfer the cells from the filter to the slide. Several slides, each containing a representative population of the exfoliated cells can be prepared from a single suspension [48]. Figure 10.29 shows a photograph of slides prepared using the conventional manner (b) and using R the ThinPrep device (a). Trials comparing conven-

tional Pap to a thin layer Pap showed that the thin layer method results in an overall 18% higher detection rate of abnormalities than conventional cytology. Based on these results, this new technology was approved by the US FDA in 1996. Further studies indicate that the use of the thin layer Pap decreases the proportion of inadequate specimens, improves the sensitivity, and reduces the specimen interpretation time compared to the conventional Pap smear [44]. There is additional cost associated with the preparation of the liquid based cytology specimen. A conventional Pap costs around $15; ThinPrep adds about $15–25 to this cost [49].

doctors may suggest a woman get screened more often if she has certain risk factors, such as HIV infection or a weakened immune system. Women 70 and older who have had three or more consecutive Pap tests in the last ten years may choose to stop cervical cancer screening. Do women follow these recommendations? Around 79% of women in the USA report having had a Pap smear in the last three years. Adherence to screening is slightly lower for women with no health insurance and for women with less than a high school education [14, 50].

Automated Pap smears Currently, Pap smears are examined by highly trained cytotechnologists. A significant amount of training is required to be able to accurately identify smears containing pre-cancerous or cancerous cells. The lack of trained personnel is a barrier to screening in many developing countries [47]. Automated cytology devices use a microscope with autofocus and a motorized, computer controlled stage coupled to a high resolution video camera (Figure 10.30). Digital images are captured and sent to a computer, where image processing algorithms are applied to interpret the images and classify the slides. Images are first segmented, to separate cells from background objects, like debris or inflammatory cells. Morphologic parameters are then calculated such as the cell size, nuclear size, the nuclear to cytoplasmic ratio and

Technologies for early detection and prevention of cancer

c Figure 10.30. An automated Pap smear machine. Courtesy and  Becton, Dickinson and Company.

the texture of chromatin within the nucleus. In addition to features that cytologists normally use, more advanced morphologic parameters can be calculated. Abnormalities can then be detected by comparing the distributions of measured cells to those of known normal and abnormal reference cases. Classification algorithms are used to combine measured parameters and make a determination of whether the specimen is normal or abnormal. Statistically based algorithms, hierarchical decision trees or neural networks are examples of types of classifiers, each of which consists of a set of rules to classify the data [48]. In 1998, the FDA approved the use of such a R device called the AutoPap Primary Screening System to sort out 25% of smears that do not require human review because they are negative. The device can scan about 200 slides per day [44]. Slides containing potential abnormalities are ranked in order of abnormality. Slides with the lowest probability of abnormality are not ranked and reported as requiring “no further review.” This approach can reduce the workload of the cytotechnologist, allowing him or her to focus on those slides most likely to contain abnormalities. In addition, the device is used to rank the 15% of slides with the greatest likelihood of abnormalities for re-review (Figure 10.31). These slides may be used instead of the 10% random selection of slides for quality control as mandated by CLIA [51]. Clinical studies of this technology have shown that R device outperformed human review by a the AutoPap factor of five to seven times when used to rescreen the 10% of negative Pap smears which must be examined


Figure 10.31. Slides obtained from a Pap smear are ranked according to abnormality before being reviewed by a cytotechnologist [44]. Reprinted with permission from Elsevier (The Lancet Oncology, 2001, Vol. 2 No. 1, pp. 27–32). R for quality control purposes [44]. The use of AutoPap adds between $3-$10 to the cost of the Pap test [52].

HPV DNA testing Cervical cancer is caused by infection with the human papillomavirus (HPV). A new test has been developed to determine whether a patient is infected with HPV. Following a Pap smear, the remaining material on the spatula can be tested to determine if HPV DNA is present. The DNAwithPapTM Test is FDA-approved for routine adjunctive screening with a Pap test for women age 30 and older [53]. HPV is found so frequently in women under the age of 30 that it is not useful to indicate risk of cervical cancer and its precursors. However, if viral infection persists after age 30, there is an association with increased risk of cervical cancer and its precursors. Clinical studies have shown that the sensitivity of DNAwithPapTM is greater than that of the Pap smear alone or a liquid Pap. The sensitivity of DNAwithPapTM is 80–90%, while the specificity is 57–89% [54]. In Europe, as of 2006, the use of HPV tests was not currently included in basic screening [55].

VIA The use of Visual Inspection with Acetic Acid (VIA) is being explored as an alternative to Pap smear and colposcopic examination in many developing countries. VIA consists of simple visual examination of the cervix with the naked eye by a trained healthcare provider before and after application of acetic acid. VIA relies on the acetowhitening of pre-cancerous lesions. It requires only


Biomedical Engineering for Global Health

HPV DNA testing: how does it work? As of 2007, Digene’s hc2 High-Risk HPV DNA Test(tm) R 2 (DNAwithPapTM ), based on Hybrid Capture technology, is the only FDA approved method for HPV DNA testing. When used for the purpose of cervical cancer screening, the FDA requires that it only be used in women over the age of 30, in conjunction with a Pap test. In most cases, HPV DNA testing may be performed on the same sample of cells collected for the Pap test. Following collection, DNA is extracted from the cell sample and denatured and single stranded RNA probes for the 13 highly oncogenic types of HPV are added. If HPV DNA is present, it hybridizes to the probe RNA. Antibodies specific to DNA–RNA hybrids capture the hybrids and bind them to the wells of a microtiter plate. The complex is then enzymatically digested resulting in the emission of light produced by a chemiluminescent substrate conjugated to the enzyme. The intensity of light indicates the presence or absence of HPV DNA in the patient’s sample [44, 53, 56]. The figure in this box is reproduced with kind permission of QIAGEN GmbH, Germany; QIAGEN copyright.

low technology equipment, and results are available in a few minutes. A recent review of the performance of

series of 1921 women screened in Peru, Jeronimo found that the VIA positivity rate dropped from 13.5% in the

VIA in nine studies involving more than 40,000 women in South Africa, India, Zimbabwe, China, and the Philippines found that VIA has similar sensitivity to that of Pap smear screening, but lower specificity, although some

first months to 4% during subsequent months of a two year study; the drop in positivity rate was hypothesized to be due to a learning curve for the evaluator [47].

studies suffered from verification bias [57]. In a study of 18,675 women in India, Sankara-


narayanan found that the sensitivity of VIA for detection of high grade squamous intraepithelial lesions (HSIL) was 60.3% and the specificity was 86.8% relative to the gold standard of colposcopic directed biopsy of colposcopically abnormal lesions [61]. The advantage of VIA is that it is an inexpensive test that does not require lab infrastructure. Providers can be trained to perform the test in five to ten days. Consumables required are cheap and universally available. Because results are available immediately, patients can be treated at the same visit. However, there are concerns that the low specificity of VIA may lead to over-diagnosis and treatment [62]. Because VIA relies on visual interpretation, defining objective criteria for a positive lesion and training operators to correctly implement these criteria are crucial. In a

The use of digital image analysis (DIA) may provide a simple solution to reduce the subjectivity and improve the specificity of VIA. Advances in consumer electronics have led to inexpensive, high dynamic range CCD cameras with excellent low light sensitivity. At the same time, advances in vision chip technology allow high quality image processing in real time. These advances may enable acquisition of digital images of the cervix in a relatively inexpensive way, with or without magnification. Moreover, automated image diagnosis algorithms based on modern image processing techniques has the potential to replace clinical expertise, which may reduce a considerable amount of the system cost. A recent pilot study showed that digital images of the cervix can be obtained using a simple and inexpensive device, and that automated image analysis

Vaccines to prevent HPV infection and cervical cancer HPV is the most common sexually transmitted disease in the USA. Today, more than 20 million people in the US harbor HPV. 80% of women will test positive for HPV by age 50. As we have seen, HPV infection usually does not cause any symptoms, but in some cases it can lead to cervical cancer. In 2006, a new vaccine to prevent HPV infection was licensed for use in girls and women aged from nine to 26 years in the USA. The vaccine, Gardasil, protects against four strains of HPV. Two of these HPV types (16, 18) are responsible for 70% of cervical cancers; combined, the 4 HPV strains covered by Gardasil account for about 90% of genital warts. At the end of Figure courtesy of CDC. 2006, Gardasil had been approved in 49 countries. Gardasil is made by inserting the gene for a protein found in the HPV capsid (L1) into a different virus or yeast. Recombinantly produced HPV capsid protein then self-assembles into virus like particles (VLPs). While these empty shells do not contain the cancer causing DNA of HPV, their shape is sufficiently similar to that of the HPV virus so that the immune system triggers a protective response against future HPV infection (the figure shows the L1 protein of HPV16 assembling into a virus like particle that does not contain HPV DNA). While Gardasil can protect against new HPV infections, it is not effective for women who have already been exposed to HPV. The length of time that patients will be protected following vaccination is currently not known. As of 2006, the vaccine had been tested in more than 3000 women who had been followed for five years. The vaccine was protective throughout this period, but it is not known whether booster shots will be required over longer periods of time. Trials of Gardasil and a second promising HPV vaccine made by Glaxo SmithKline (GSK) are currently underway in more than 50,000 subjects. Currently, Gardasil is given as a series of three shots over a six month period; the cost of the vaccine is $360. This cost is a barrier even in developed countries, and is likely to limit its immediate impact in developing countries. For example, the HBV vaccine was licensed in 1981 in industrialized countries, but it took 10–15 years for it to be used in wealthier developing countries and over 20 years before children in the poorest countries had access to the vaccine [60]. Developing countries may also face difficulties in providing widespread access to a vaccine that is targeted towards girls and young women. Vaccines for adolescents are often given through school programs, but girls in developing countries are less likely to be in school than boys. Gender specific immunization may be culturally unacceptable in some settings. Many vaccination programs have been damaged due to rumors that vaccination is a plot to sterilize girls [60]. The stigma associated with a vaccine targeted against an STD may exacerbate such rumors. Such rumors derailed polio eradication campaigns in Nigeria and India, resulting in global consequences. Will the HPV vaccine eliminate the need for cervical cancer screening? Currently available vaccines do not protect against all types of HPV that cause cervical cancer, so women who receive the HPV vaccine will still need to be screened for cervical cancer. Additionally, if women don’t get all three doses of the vaccine or if they have already been exposed to HPV prior to being vaccinated they may not be protected [58–60]. Figure credit: X. Bosch and N. Munoz/IARC, IBSCCS, and Multicentric Studies (N = 3045). From Science 29 April 2005: Vol. 308. no. 5722, pp. 618–621. Reprinted with permission from AAAS.


Biomedical Engineering for Global Health

Part II: Peace Corps and rice visits: July 2–July 4, 2007 Tessa Swaziland The next day, Tandi took me to the NCP (Neighborhood Care Point). There were several in the community, but this one in particular was also the kagogo, which is the central meeting place for the community. I met the secretary of the kagogo, and he asked me lots of questions about Baylor. No one in the community knew about the clinic, and he was curious as to who was eligible to go there, if it was free, and how they could become an outreach site. Most of the people wouldn’t be able to afford the 42 rand (US $6) it would take to go to and from the clinic, so ideally, he would get Baylor to come to the community. The NCPs are where orphans and vulnerable children can come for the day to receive meals and a bit of education. Tandi said that the community was pretty good about taking in the children but struggled to support them. Then NCPs filled this gap and provided as much support as possible, although often, it isn’t enough either. Talia (a Canadian who visited with Rachel and Lindsay) worked a lot with orphans in Botswana. One of the services her NGO provided was gift baskets for the orphans. Once families learned about this, they started taking in as many orphans as they could in order to receive the baskets, which they would then sell. The orphans remained just as abandoned and starving as before. Tandi said that this wasn’t really a problem in her community, but there were many others. For example, an orphan could go to school if they could prove (with death certificates) that they were indeed orphans. This is virtually impossible for many reasons. Many of them never knew their fathers, who left the mother when she was pregnant. Even if they knew both of their parents, no one gets a death certificate unless they go to a city far away and deal with some complicated legal procedure. So, unless there is someone who cares enough for the child to deal with the hassle and who is wealthy enough to afford it, there is no way for the orphan to prove that they lack parents. Thus, all they are left with are the NCPs. I listened for a bit as Tandi and the secretary discussed some of their projects – a community garden, fundraising for NCP renovations (most of them were dirty, stick-in-the-mud structures), education campaigns, and other events. During this conversation, I discovered that children become sexually active as early as twelve. I also learned that men fear the HIV stigma more than women (probably because it might limit the number of girlfriends they could have), while women were much more open and willing to address the problem. He told us that there was an article in the paper about a doctor who was telling many of his patients that they weren’t actually HIV-positive even though they’d been told at a VCT (volunteer counseling and testing) clinic that they were. He thought they were lying about it because they were afraid that they would lose their jobs if the HIV rate dropped and funding for HIV/AIDS programs dropped. Tandi responded that it is much more likely that the one doctor was lying than everyone at the VCTs, and in addition, many people try to place blame elsewhere in order to avoid taking responsibility for their actions (which caused them to get the disease). After that, I visited the school and nearby clinic. They were pretty much what I expected – about the same as the Vuvulane clinic, and the school was much like the school I worked at in Nicaragua. We looked at the picture of the map Tandi was painting with her class and I took photos of some of the HIV/AIDS awareness signs. Realizing that her watch had stopped, we rushed off to catch a combie and make our way back to Manzini. At that point, we split up. She headed over to her friend’s community to help with a workshop, and I headed back to the COE in Mbabane.

Technologies for early detection and prevention of cancer


Figure 10.32. Many young women develop an HPV infection during adolescence or young adulthood. In some women, HPV infection leads to pre-cancerous changes in the cervix. Regular Pap tests can identify these pre-cancerous changes, allowing treatment before cervical cancer develops. In the future, the availability of an HPV vaccine may reduce the incidence of HPV infection and reduce the frequency with which screening is needed. Alternative methods of screening, such as the HPV DNA test, may improve the sensitivity of screening. Schiffman, Castle. The promise of Global Cervical Cancer Prevention. c 2005 NEJM. 353: 2101–04. Copyright  2005 Massachusetts Medical Society. All rights reserved.

algorithms correctly identify histologically neoplastic tissue areas with a sensitivity of 79% and a specificity of 88% [63]. In summary, although cervical cancer is a completely preventable disease, it is the third leading cause of cancer death in women in the world [6]. Cervical cancer is caused by infection with HPV. HPV infection can initiate a transformation that results in a pre-cancerous lesion. If we detect and treat these common pre-cancerous lesions, we can prevent the development of cervical cancer. Current screening and detection using the Pap smear followed by colposcopy and biopsy has been proven to reduce both the incidence and mortality of cervical cancer. However, we have insufficient resources to screen using these technologies in developing countries. New technologies, such as automated reading of Pap smears, HPV testing, visual inspection, and digital image analysis (VIA and DIA) technologies may provide the improvements in performance at a sufficiently low cost to enable screening in resource poor settings where the vast majority of cervical cancer occur. Coupled with vaccines to prevent HPV infection, these tech-

nologies have the potential to reduce both the incidence and mortality of cervical cancer (Figure 10.32).

Prostate cancer As we have seen, prostate cancer is the most common cancer diagnosed in men in the USA, with 218,890 new cases diagnosed annually. Prostate cancer is the second leading cause of cancer death in men, causing more than 27,050 deaths each year in the United States [1]. Worldwide, more than 679,023 new cases of prostate cancer are detected annually, making prostate cancer the second most common cancer in men [6]. Figure 10.33 shows the incidence and mortality rates of prostate cancer throughout the world. Risk factors for development of prostate cancer include advanced age, race (incidence rates are one and a half times higher in African Americans), and a family history of prostate cancer [1]. Figure 10.39 shows the location of the prostate gland. The prostate gland contributes enzymes, nutrients and other secretions to semen. Figure 10.34a shows a photograph of the normal prostate, while Figure 10.34b shows


Biomedical Engineering for Global Health

Figure 10.33. The incidence of prostate cancer is highest in North America. Mortality rates of prostate cancer are substantially lower than incidence rates. Parkin, Bray, Ferlay, Pisani. Global Cancer Statistics. 2002. CA Cancer Journal for Clinicians, 55: 74–108.

histologically stained sections of normal prostate tissue. The normal prostate consists of several branched glands leading to the urethra. These glands are covered by a single layer of columnar epithelial cells. In the normal prostate, the nuclei of these cells occupy approximately 1/4 of the cell area (Figure 10.35). However, in pre-cancerous lesions, the nuclei of these epithelial cells become substantially enlarged (Figure 10.36) and multiple layers of cells stack atop one another. As these cells invade beneath the basement membrane lining the

Prostate cancer is a slow but continuously growing cancer. Generally, pre-clinical asymptomatic forms of the disease can develop as early as age 30. This disease can remain latent for up to 20 years. In some patients, pre-cancerous lesions can progress to aggressive, malig-

ducts, invasive prostate cancer develops. Initially, the lesion is localized to the prostate; at the microscopic level, the cancerous epithelial cells are found throughout the entire prostate (Figures 10.37a,b).

ing age. Prostate cancer is often asymptomatic in the early stages. When present, the signs and symptoms of prostate cancer include weak or interrupted urine flow or the inability to urinate. These symptoms are the same

nant cancer. The peak incidence of prostate cancer occurs in the seventh decade of life [11]. Figure 10.38 shows the risk of developing prostate cancer in the next five years as a function of a patient’s current age; the risk rises dramatically with increas-

Technologies for early detection and prevention of cancer


Serum PSA test Prostate specific antigen is a glycoprotein with a molecular weight of 34,000 Dalton. It is responsible for liquefaction of semen. PSA is highly specific for prostate tissue; it was first discovered when scientists were searching for a potential marker that could be used in investigation of rape crimes. The PSA test is a blood test to measure levels of PSA in the serum. In the test, a serum sample is added to a tube containing two types of anti-PSA antibodies that recognize different antigenic sites on the PSA. One type of anti-PSA antibody is conjugated to an enzyme called alkaline phosphatase; the second type of anti-PSA antibody is conjugated to paramagnetic particles. If PSA is present in the sample it binds to both antibodies forming a sandwich complex. A magnetic field is applied to separate the magnetic particles. The sample is washed to remove unbound alkaline phosphatase conjugate; thus, the remaining alkaline phosphatase is proportional to the amount of PSA present in the sample. A chemiluminescent substrate called Lumigen PPD is added. Alkaline phosphatase causes cleavage of the phosphate group on the Lumigen PPD producing an intermediate product. The intermediate product decomposes and generates chemiluminescence; this signal decays with a half life of several minutes. The light intensity produced is a direct measure of enzyme present. O O OCH 3

The serum PSA test was first approved by the FDA in 1986 to monitor patients who had been treated for

O O OCH 3 AP alkaline buffer

prostate cancer to determine whether they had a recurrence of disease. In the early 1990s physicians began to use the test to screen patients who were at risk for developing prostate cancer. Two large studies


OPO3Na2 Lumigen PPD O



OCH3 light

+ O−

have been carried out to study the accuracy of the PSA test. When the cut-off value for an abnormal PSA test is

set at 4 ng/l, its sensitivity has been reported to be 44–46%, with a specificity of 91–94%. The cost of a PSA test is approximately $50 [65–70]. A number of approaches have been suggested to improve the sensitivity and the specificity of PSA based

Courtesy Lumigen, Inc.

screening. Adjust cut-offs with age since PSA levels increase with age. Adjust cut-offs with ethnicity since African American males tend to have higher PSA values. Monitor annual increases in PSA levels rather than absolute values. Adjust PSA levels by the size of the prostate (PSA density). Measure the fraction of free PSA relative to that bound to plasma proteins.

as those of prostate enlargement, thus are not diagnostic [1]. Prostate cancer is treated with a combination of

where the five year survival rate is 100%. When disease has metastasized to distant organs, the five year survival rate of prostate cancer is only 33.3% [1]. Thus,

surgery, radiation therapy, hormone therapy, and chemotherapy [64]. In the United States, the five year survival rate for all stages of prostate cancer combined is quite high at 99.9%. This is due to the effectiveness of treatments for cancer which is localized to the prostate,

early detection of prostate cancer is important. There are two tests which have been widely used to screen the general male population for prostate cancer, although there is considerable controversy regarding the most appropriate use of these tests. The first test is a


Biomedical Engineering for Global Health



Figure 10.35. A slide showing a normal prostate; note that the nuclei of normal cells take up roughly a quarter of the cell area. Photo courtesy of Laura P. Hale, M.D. PhD., Duke University Medical Center.

Figure 10.34. (a) Photograph of the normal prostate; 2004 Bostwick Laboratories, Inc. (b) Histologically stained section of a normal prostate. Reprinted with permission from Andrei Gunin, the Department of Obstetrics and Gynecology Medical School Chuvash State University.

Figure 10.36. A slide showing a pre-cancerous prostate gland, in which the nuclei of the cells have become enlarged. Dharam M. Ramnani, M.D.;

simple blood test to measure levels of a protein called prostate specific antigen (PSA). PSA is a protein found on the surface of epithelial cells in the prostate. When

gland lies close to the rectum, and its size can be felt by placing a gloved finger inside the rectum. Prostate enlargement can be a sign of either prostate cancer

prostate cancer develops, the number of epithelial cells increases and the amount of PSA found in the blood increases. A blood test can measure the levels of serum PSA quantitatively. However, other conditions which cause an increase in the number of prostate epithelial

or benign prostate enlargement [64]. Screening using the PSA and DRE tests has become one of the most commonly used cancer screening tests. More than a half of men over age 50 report having a recent serum PSA test and a digital rectal examination, although these figures

cells, such as benign enlargement of the prostate, can also cause PSA levels to be elevated [64]. The second test is to palpate the size of the prostate gland in a procedure called a digital rectal exam (DRE). The prostate

drop to less than 33% for men without health insurance [14]. If screening tests for prostate cancer are positive, further diagnostic tests to confirm or exclude the presence

Technologies for early detection and prevention of cancer




Figure 10.38. As the graph above shows, the risk of developing prostate cancer in the next five years increases dramatically with age. BC Cancer Agency. PSA Screening information for patients, May 2009. Available at 375628D8-AB6F-4523-BFD4-C854FDA705F8/4510/ PSAwebBrochure.pdf.

Figure 10.37. Invasive prostate cancer at the (a) macroscopic and (b) microscopic levels. Courtesy of the University of Washington, Department of Pathology, NCI/Otis Brawley.

of prostate cancer are required. To confirm the presence of cancer, physicians obtain small pieces of prostate tissue called core needle biopsies. Biopsies are then sectioned, stained and observed under a light microscope to examine the epithelial cells of the prostate. A biopsy of prostate tissue is obtained by inserting a needle through the wall of rectum into the prostate (Figure 10.39). A positive screening test does not indicate where in the prostate a lesion might exist, so multiple biopsies are performed. Typically at least ten core biopsies are obtained to sufficiently sample the prostate tissue; the procedure is performed with local anesthetic. The precise positioning of the needle is guided by ultrasound imaging. Small fragments of the prostate are

Figure 10.39. A diagram showing the prostate biopsy procedure.

then removed from the needle, processed, and examined under a microscope [71]. The cost of obtaining and processing a prostate biopsy is approximately $700 [70]. If prostate cancer is detected when it is still localized to the prostate, physicians generally recommend one of two courses of action. The first is radical prostatectomy, a surgical procedure to remove the prostate. While this procedure is usually curative, because it removes the cancerous cells, it has some very serious side effects. Because important nerves which control


Biomedical Engineering for Global Health

Table 10.12. The ten year survival rates for three grades of prostate cancer following either surgery or conservative treatment. The earlier the cancer is detected, the less difference between survival rates for the two treatments. Cancer grade

Surgery ten year survival

Conservative ten year survival

Grade I



Grade II



Grade III



Source: Lu-Yao, GL and Yao, SL. Population-based study of long-term survival of patients with clinically localised prostate cancer. Lancet 1997; 349; 906–910.

as 20–50% of men who have died with no symptoms of prostate cancer have been found to have prostate cancer at autopsy [69]. Since the treatment of prostate cancer has significant side effects, patients and physicians are faced with difficult decisions about whether to treat the disease or watch the disease. Thus, the question of whether to screen the general male population for prostate cancer has a complicated answer. Localized prostate cancer is curable, and advanced prostate cancer is fatal, indicating the benefits of detecting disease early through screening. While screening clearly has potential benefits, it also has potential risks. A positive screening results leads

bladder function and sexual function are located in the same area as the prostate, they can be damaged during

to a prostate biopsy, an expensive, reasonably invasive and uncomfortable procedure. For those patients whose

this surgical procedure [64]. Following radical prostatectomy, between 2% and 10% of men experience incontinence, and between 30% and 90% of men experience impotence [11]. Because of the seriousness of these side

screening test is falsely positive, this biopsy is unnecessary. Furthermore, because prostate cancer is such a slow growing cancer found in older men, screening may lead to over-detection of latent cancers. If we screen,

effects, other physicians recommend more conservative management of prostate cancer, choosing to watch the

we may detect many cancers that would never have produced symptoms before the patients died of other

patient until symptoms develop, and then offering treatment [64]. Because prostate cancer is a relatively slow growing cancer, there is some controversy over whether detec-

causes. Let’s examine some of this clinical evidence regarding the efficacy of screening. In Tyrol, Austria, the mortality from prostate cancer was constant from 1970 to

tion of very early disease makes a difference in patient outcomes. Localized prostate cancer is classified into

1993, prior to the introduction of mass screening. In 1993, mass screening for prostate cancer using digital

three grades based on the severity of the disease. A study to examine the ten year survival rates for localized prostate cancer found that the survival rates for surgery and conservative therapy were nearly the same for grade

rectal examination and serum PSA began [74]. Between 1993 and 2000, the mortality associated with prostate cancer decreased 44% in Tyrol [75]. While this study was not designed with a control group, cancer mortal-

I disease, but were substantially higher when grade II or grade III disease were treated surgically (Table 10.12) [72]. This illustrates one of the challenges of screening for prostate cancer. Prostate cancer is a slow growing cancer; the average patient does not show symptoms for an average of ten years following the initial development of prostate cancer. Because prostate cancer occurs later in life, most men with prostate cancer actually die of other causes. For example, a 50 year old man has a 42%

ity remained constant in other parts of Austria where screening was not performed. Other, more carefully designed and controlled studies have shown contrasting results. One completed randomized clinical trial of digital rectal examination and PSA to screen found no difference in the number of prostate cancer deaths between groups randomized to screening and usual care [76]. One prospective clinical trial in Canada suggested that screening with PSA could reduce the mortality due to prostate cancer by 67%,

chance of developing microscopic prostate cancer sometime in his life, a 10% chance of having this cancer diagnosed, but only a 3% chance of dying of it [73]. As many

although the study was widely criticized for design and analysis flaws. Two large randomized clinical trials of screening are underway – the European Randomized

Technologies for early detection and prevention of cancer


Costs of screening for prostate cancer We can examine the predictive value of the PSA screening test and the cost to find prostate cancer with this test. Let us assume that we test one million men between the ages of 50 and 59 for prostate cancer using a serum PSA test with a sensitivity of 44% and a specificity of 91%. The expected prevalence of prostate cancer is 10% in this population. The cost to screen is $50/patient, and a high serum PSA results in a follow up biopsy which costs $700. What are the positive and negative predictive values in this situation? What is the cost to screen the entire population? What is the cost to biopsy all men with positive tests? What is the cost/cancer found? To answer these questions, we fill in the 2 × 2 table shown here.

Disease Present Disease Absent

Test Positive 44,000 81,000 # Test Pos = 125,000

Test Negative 56,000 819,000 # Test Neg = 875,000

# with Disease = 100,000 # without Disease = 900,000 Total Tested = 1,000,000

The PPV and NPV are then PPV = 44,000/125,000 = 35% and NPV = 819,000/875,000 = 94%. Thus, a man with a negative PSA test has a 94% chance of not having prostate cancer. However, a man with a positive PSA test only has a 35% chance of having prostate cancer. The cost to screen the entire population is $50 million dollars. In addition 125,000 men will have a positive PSA test and require a biopsy. Note that 81,000 of these biopsies are unnecessary! The cost to biopsy this group is 81,000∗ $700 = $56,700,000. Using this strategy we will find 44,000 cancers at a cost per cancer found of $56,700,000/44,000 = $1288 [24, 69, 70]. Are the costs of screening with the PSA test a good use of healthcare resources? In Chapter 11, we will examine how to calculate the cost effectiveness of different health interventions.

study of Screening for Prostate Cancer (ERSPC) involving 200,000 men, and the Prostate, Lung, Colorectal,

screened. However, screened patients do survive for a longer period following diagnosis of their cancer, only

Ovarian cancer (PLCO) study involving 74,000 men in the USA – but results are not expected to be available for years [77].

because their cancer was detected before it produced clinical symptoms. This apparent increase in survival time following diagnosis is called “lead time bias,” indi-

The fact that prostate cancer is such a slowly growing cancer makes it difficult to perform a controlled experiment and test whether an intervention truly reduces mortality. Figure 10.40 shows the natural history of prostate cancer vs. time. Once a microscopic cancer develops, it typically takes ten years before symptoms

cating that the new intervention simply lead to earlier diagnosis without truly changing the outcome. Thus, randomized clinical trials must be carefully designed to minimize lead time bias [78]. Given the limited clinical evidence currently available, different countries approach prostate cancer

develop which would lead to a diagnosis even without the use of any screening tests. In this scenario, a typical patient survives 15 years beyond the initial diagnosis [69]. How does this sequence of events change if we screen for early disease? By screening asymp-

screening in different ways. In the United States, there are conflicting recommendations regarding screening (Table 10.13). The American Cancer Society recommends men aged 50 or older with more than a ten year life expectancy should be screened with DRE and PSA.

tomatic patients, we detect disease earlier, as much as ten years before symptoms develop. If our ability to detect prostate cancer early does not change the natural history of the disease, these screened patients do not live to be any older than patients who have not been

The American College of Preventive Medicine recommends that men aged 50 or older with greater than ten year life expectancy should be informed of the potential benefits and risks of screening and make their own decision [69]. The US Public Service Task Force in


Biomedical Engineering for Global Health

Figure 10.40. The natural history of prostate cancer versus time. The apparent increase in survival associated with screening is called lead time bias.

their Guide to Clinical Preventive Services recommends against screening using DRE or serum PSA [24]. While they find good evidence that PSA screening can detect early stage prostate cancer, they conclude that there is mixed and inconclusive evidence that early detection improves health outcomes. They note that screening is associated with important harms, including frequent false positive results and unnecessary anxiety, biopsies, and potential complications of treatment of some cancers that may never have affected a patient’s health. They conclude that the available evidence is insufficient to determine whether the benefits outweigh the harms for a screened population. In Europe, screening is not currently recommended, because it is not believed that there is sufficient evidence to indicate that screening reduces mortality [69]. How has the incidence and mortality of prostate cancer changed in the USA following the introduction of screening? Routine screening has led to a dramatic increase in the number of cases of prostate cancer detected (Figure 10.41). The increase in incidence was accompanied by a shift in the stage at which prostate cancer is detected to earlier clinical stages. Over

the period 1950–1996, the incidence of prostate cancer increased by 190% in the USA (Table 10.14). Over this same period, the five year survival rate increased from 43% to 93%. However, the increase in five year survival rate may simply reflect the lead time bias associated with earlier detection. Over this same period, the mortality of prostate cancer in the USA has actually increased by 10%. In contrast, during this same period cervical cancer screening led to a 79% decrease in the incidence of cervical cancer and a 76% reduction in the mortality of cervical cancer. The next type of cancer we will consider is ovarian cancer; in contrast to cervical cancer and prostate cancer where screening tests are available, there is currently no good screening test for ovarian cancer. Table 10.14 shows that the incidence and mortality of ovarian cancer have not changed appreciably from 1950–1996 [79].

Ovarian cancer We have considered two cancers where screening tests are available: cervical cancer and prostate cancer. In our final example, we turn to a cancer where there is no

Technologies for early detection and prevention of cancer Table 10.13. There is no consensus on how, if, or at what age men should begin being screened for prostate cancer in the USA. Modified from cancer/HQ01273. With permission from Mayo Foundation for Medical Education and Research. All rights reserved. Organization


American Urological Association

Men over 50 should consider testing. Men at high risk should begin testing at age 45.

American Cancer Society

Offer the PSA and DRE tests annually beginning at age 50 to men who have a ten year life expectancy and to younger men at higher risk


Cancer Incidence Rates* Among Men, US, 1975–2004 Rate Per 100, 000 250 Prostate

200 150


Evidence is insufficient to determine whether the benefits of screening outweigh the harms.

American Academy of Family Medicine

Physicians should counsel men between ages of 50 and 65 about known risks and uncertain benefits of screening so they may make an informed choice.

American College of Physicians

Colon and rectum

50 0

1975 1978 1981

Melanoma of the skin

1984 1987 1990 1993 1996 1999 2002

*Age-adjusted to the 2000 US standard population and adjusted for delays in reporting. Source: Surveillance, Epidemiology, and End Results Program, Delay-adjusted Incidence database: SEER Incidence Delay-adjusted Rates. 9 Registries, 1975–2004, National Cancer Institute, 2007.

Figure 10.41. Cancer incidence rates for men in the US vs. time. A large increase in prostate cancer incidence was reported shortly after initiation of PSA based screening in the late 1980s to early 1990s [21]. Reprinted with permission from the American Cancer Society, Inc. from All rights reserved.

22,430 new cases of ovarian cancer, representing 3.3% of all cancers in women. It was estimated that 15,280 women would die as a result of ovarian cancer in 2007 in the USA [1]. Worldwide there were 190,000 new cases of ovarian cancer and 114,000 deaths in this same year. The highest rates

Physicians should describe potential benefits and known harms of screening, diagnosis and treatment, listen to patients’ concerns and individualize the decision of whether to screen

of ovarian cancer occur in Scandinavia, Eastern Europe, USA, and Canada (Figure 10.43) [11].

adequate screening test – ovarian cancer. The ovaries are part of the female reproductive system (Figure 10.42a,b) and are located adjacent to the fallopian tubes. In the USA in 2007, there will have been an estimated

The treatment for ovarian cancer involves surgery, and for advanced disease involves radiation therapy and chemotherapy. The five year survival rate for all stages of ovarian cancer is 45% [1]. There are four stages of ovarian cancer; when detected early, the five year

Table 10.14. Changes in survival rates and incidence for several cancer types since 1950 [79]. Five year survival, %

Urinary bladder

Non-Hodgkin lymphoma

Routine screening is not recommended because there is not consensus on whether screening and early treatment reduces mortality.

US Preventive Services Task Force

Lung & bronchus


% Change (1950–1996)



Absolute increase in five year survival, %





















Source: JAMA 2000, 283: 2975. Copyright 2000 American Medical Association.


Biomedical Engineering for Global Health


survival rates are much higher. Some 90% of women diagnosed with stage I ovarian cancer, when the disease is localized to the ovaries, survive five years beyond their initial diagnosis. However, the five year survival rate for metastatic, stage III–IV ovarian cancer is only 25–37% [80]. Unfortunately, because of the lack of good screening tests and the fact that early ovarian cancer produces relatively few symptoms, more than 70% of women diagnosed with ovarian cancer are diagnosed at stages III and IV [1]. Table 10.15 compares the ratio of mortality rate to the incidence rate for the ten most


common cancers in women; ovarian cancer has one of the highest mortality to incidence ratios, second only to pancreatic cancer and lung cancer [81]. On average, women who die of ovarian cancer lose 18 years of life to the disease [82]. Ovarian cancer is said to “whisper” because the symptoms are so vague. Symptoms can include unexplained changes in bowel and/or bladder habits; gastrointestinal upset; unexplained weight loss or weight gain; pelvic and/or abdominal pain, discomfort, bloating or swelling; a constant feeling of fullness; fatigue; abnormal or postmenopausal bleeding and pain during intercourse. Frequently, women (and their physicians) will

Figure 10.42 (a) Diagram of ovary indicating stages of ovulation. (b) Histologic photograph of ovary.

attribute these symptoms to those normally experienced with aging.

Figure 10.43. Global incidence rates of ovarian cancer.

Technologies for early detection and prevention of cancer


Table 10.15. Mortality/incidence ratio for ten most common solid cancers in women in the USA.

From Rosenthal et al., Clinical Obstet. and Gynecol., 49(3)433–447, 2006.

There are a number of factors that put a woman at higher risk for developing ovarian cancer. The most important risk factors are a personal or family history of breast, ovarian, endometrial, prostate or colon cancer, particularly having one or more first-degree relatives (mother, sister, daughter) who have ovarian cancer. Ovarian cancer is sometimes associated with a mutation in the BRCA1 or BRCA2 gene. Hereditary ovarian cancer accounts for about 10% of cases [80]. In addition, the risk of ovarian cancer increases with the more lifetime cycles of ovulation that a woman has undergone. Thus, women who have undergone hormonal treatment for infertility, never used birth control pills, and who never became pregnant are at higher risk for ovarian cancer. In addition, the use of high dose estrogen for long periods without progesterone may also increase the risk of

or more follicles undergoes a transformation in preparation for ovulation. A primordial follicle enlarges and develops into a primary follicle. The follicle continues to enlarge and move toward the surface of the ovary. The secondary follicle then merges with the ovarian surface, ruptures and releases the oocyte. The defect in the ovarian surface must then repair itself. The scar left behind is known as a corpus albicans. Thus, the surface of the ovarian epithelium is constantly undergoing damage and repair. During this process, epithelial cells can become transformed and lead to ovarian cancer. As the frequency of this repair process increases, so do the chances that an ovarian epithelial cell will become transformed leading to an ovarian cancer. This probably explains why the use of oral contraceptives,

developing ovarian cancer. The ovary is an almond-shaped organ that contains all the eggs that will be released over a woman’s reproductive lifetime (Figure 10.42a,b). The ovary is lined by a single layer of epithelial cells. Beneath the epithelium, the ovary contains spherical follicles, each containing a single oocyte (egg), in a region known as the ovarian cortex. At the very center of the ovary, blood

pregnancy and breast feeding reduce the risk of ovarian cancer development. Because ovarian cancer does not generally produce symptoms until very advanced stages, there has been substantial research to develop good early detection tools. Three are available, but all suffer from significant limitations; these techniques include: (1) pelvic and rectal examinations, (2) the CA-125 blood test, and (3) transvaginal ultrasound.

vessels bring in oxygenated blood and nutrients in a region known as the ovarian medulla. Each month, one

Pelvic and rectal examinations are normally conducted when a woman has a Pap smear. In this


Biomedical Engineering for Global Health

procedure, a physician manipulates the abdomen to feel the uterus and ovaries to find abnormality in shape or size. While this procedure can detect large changes asso-


ciated with advanced ovarian cancer, it is unlikely to detect early stage ovarian cancer. The CA-125 blood test is similar to the use of PSA to screen for prostate cancer. Ovarian cancer cells produce a protein called CA-125 which is released into the bloodstream; 80% of women with advanced ovarian cancer have elevated CA-125 levels [83]. In fact, physicians routinely used blood levels of CA-125 to monitor women following treatment for ovarian cancer – it is a sensitive indicator of persistent or recurrent disease [84]. Unfortunately, CA-125 levels are very unreliable for detecting early cancer, particularly in pre-menopausal women. The reasons are two fold: first, CA-125 lev-


els are often not elevated in early ovarian cancer. Second, CA-125 levels can be elevated by conditions such as pregnancy, endometriosis, uterine fibroids, liver disease, and benign ovarian cysts [85]. Thus, in a premenopausal woman, an elevated CA-125 level is much more likely due to a benign cause than due to ovarian cancer [81]. The sensitivity and specificity of serum CA125 levels in one large Norwegian study were an overall sensitivity of 30–35%, with a specificity of 95.4% [86, 87]. In general, the sensitivity is lower for detecting early stage disease.


Improved CA-125 tests Recent attempts to improve the performance of screening using CA-125 have focused on using an algorithm that incorporates patient age, absolute levels of CA-125 and the rate of change of CA-125. Using this approach, a sensitivity of 83% and a specificity of 99.7% have been achieved [81].

Finally, ultrasound imaging can be used to visualize the ovaries. It is difficult to use ultrasound to visualize the ovaries through the abdominal wall. In order to view the small ovaries, an ultrasound probe is inserted into the vagina, and placed close to the ovaries. Using high frequency sound a picture of the ovaries is created (Figure 10.44a,b,c). Transvaginal ultrasound can detect

Figure 10.44 Transvaginal ultrasound can be used to image the c 2008 Nucleus Medical Art, All ovary. (a) Illustration Copyright  rights reserved. (b) From Richard S. Legro, MD. Diagnostic criteria in polycystic ovary syndrome, Semin. Reprod. Med. 21(3): 267–275, 2003. Reprinted by permission. (c) Reprinted with permission from Samuel Marcus MD, Medical Director.

Technologies for early detection and prevention of cancer ovarian malignancies in asymptomatic women, based on the increase in ovarian volume, and the presence of complex cysts within the ovary [88]. However, it has


Laparoscopic Procedure

poor accuracy in detecting early stage disease. A recent large study of transvaginal ultrasound to screen 14,469 asymptomatic women achieved a sensitivity of 81% and a specificity of 98.9% for the detection of ovarian cancer [89]. The only way to confirm a positive screening test for ovarian cancer is to perform a biopsy of the ovary. Because the ovaries are located in the abdominal cav-


ity, this procedure involves surgical exploration of the abdomen to visualize and potentially biopsy the ovaries. Typically, this surgery is performed through a laparoscope (Figure 10.45). As we will see in detail in Chapter 14, in this procedure a small trochar is punched through the abdominal wall and the abdomen is inflated with CO2 gas. Then, fiber optic laparoscopes are inserted through the abdomen to view the ovaries; small biopsy



Gas Filled Area Fallopian Tube Ovary


forceps can also be inserted to sample the tissue; a diagnosis of ovarian cancer can be definitely made by examining the biopsy in the same way that a cervical biopsy is examined. Approximately 1% of women undergoing laparoscopy will have a complication that will require an open surgical procedure [90]. Let’s consider what happens when we screen a group of women for ovarian cancer using the available screening and diagnostic tests. If we screen 1,000,000 women in a setting with a 0.03% prevalence of undiagnosed ovarian cancer, there are a total of 300 cases that we can possibly detect [91]. Let’s assume we use the CA125 blood test to screen our patients, and recommend that those women with an elevated CA-125 have a laparoscopy. The sensitivity of CA-125 is 35% and the specificity is 95.4% [86]. The test costs about $60 to perform [92]. In this scenario, we will spend $60 million to screen our population; the screening test will identify 105 true positives, and a staggering 45,986 false positives, all of whom will undergo laparoscopy and biopsy, which is our gold standard. The cost of laparoscopy is approximately $1500 and 1% of women undergoing laparoscopy will suffer a serious complication requiring open surgery [90, 93]. In this scenario, we will spend $1,229,871 for each cancer that we find.

Figure 10.45. (a) An ovarian biopsy is obtained during laparoscopy. Courtesy of Allon Health Center, Center for Women’s Medicine. (b) A fiber optic catheter is used to visualize the ovaries and guide biopsy direction. Reprinted with permission. Courtesy of John P.A. George, M.D., Gynecologic Endoscopy, Washington Hospital Center, Washington, DC.

Although we find only 105 cancers, 195 cancers will go undetected and 45,986 women will undergo an unnecessary laparoscopy and 460 women will suffer a complication as a result. In this scenario, the number of patients who suffer a serious complication caused by screening (460 women) exceeds the number of women correctly diagnosed with ovarian cancer (105 women).


Biomedical Engineering for Global Health

The PPV of this screening strategy is only 0.23%, the NPV is 99.98% If we use transvaginal ultrasound to screen our

exceeded a threshold level, transvaginal ultrasound was performed. In the screening group of 10,958 women, 468 women underwent 781 ultrasound exams because

same population, the outcomes improve somewhat. The sensitivity of transvaginal ultrasound is 81% and its specificity is 98% [91]. The cost to perform this imaging procedure is approximately $200 [92]. In this scenario, we identify 243 of the 300 ovarian cancers, but 19,994 false positives lead to unnecessary laparoscopies, resulting in 202 complications. While the cost to detect a case of ovarian cancer is reduced to $947,965 in this strat-

their CA-125 levels were elevated. Twenty-nine women underwent biopsy to detect six cancers. Thus, the overall positive predictive value of multi-modal screening was 6/29 = 20.7%. While the predictive value of this approach is higher, there are concerns that the sensitivity of this approach is not high enough. Despite the screening provided in this study, an additional ten women in the screening group developed ovarian can-

egy, the associated PPV is still dismally low at 1.2%; the NPV is 99.99%. We can examine the use of transvaginal ultrasound in a population with a higher prevalence of ovarian cancer. If we screen post-menopausal women over the age of

cer during a follow up of eight years. Five of the 16 cancers discovered in the screening group were stage I or II, whereas only two of the 20 cancers discovered in the control group were stage I or II. Because ovarian cancer is such a devastating disease,

45, the prevalence of undiagnosed ovarian cancer rises

there are a number of ongoing trials testing new screen-

to approximately 0.2% [94]. In our cohort of 1,000,000 women, there are 2000 cases of ovarian cancer. In this population, transvaginal ultrasound correctly identifies 1620 women with ovarian cancer. The approach results

ing approaches [81]. In the UK, a trial of 200,000 postmenopausal women is underway, comparing annual screening with CA-125 or transvaginal ultrasound to no screening [96]. Results are expected in 2012. In the USA,

in 19,960 false positives, and 216 serious complications. The cost to detect a single case of ovarian cancer is reduced to $143,438, and the PPV is 7.51%.

a trial of 78,000 is underway comparing the ability of annual serum CA-125 and transvaginal ultrasound to no screening [97]. Results of these clinical trials will help

In this population, how high does the specificity of our screening test need to be in order to achieve a PPV of 10%? A simple calculation shows that the test specificity must reach 99.9% in order to achieve even a mod-

determine future screening recommendations throughout the world.

est PPV, where one in every ten follow up laparoscopies will identify an ovarian cancer. This illustrates the difficulty of screening for a rare disease – in general, unless the specificity of the test is extremely high, the number of false positive results will far exceed the number of true positive results. If the follow up test carries any risk, then screening for a rare disease can actually cause greater harm than good.

New screening tests for ovarian cancer

Let’s examine what has happened in an actual clinical trial of these technologies to screen for ovarian cancer. The most successful results have been obtained using a combination of approaches to screen for ovarian cancer. A randomized clinical trial of 22,000 women compared no screening to a combination of screening with CA125 and transvaginal ultrasound [95]. In the group of women who were screened, CA-125 blood tests were performed annually for three years. If the CA-125 levels

Because of the limitations of current screening tests, researchers are searching for additional markers that might be useful for ovarian cancer screening. Most current cancer screening tests look for a single protein in the serum (e.g. CA-125, PSA). However, serum contains many proteins; it may be possible to identify complex patterns of serum proteins which are predictive of cancer. This field is called proteomics. In this approach, researchers use techniques to analyze the patterns made by all proteins in the blood, without even knowing what they are. The technique used to measure the pattern of serum proteins is known as mass spectrometry. In this technique, serum proteins are extracted, and bombarded with an electron beam. The electron beam has sufficient energy to fragment the proteins. This process produces

Technologies for early detection and prevention of cancer


while the strength of the signal plotted on the y-axis is proportional to the amount of protein fragment with that mass in the sample. If one does mass spectrometry using a chemically pure sample, the mass of each fragment of the molecule enables one to determine the chemical structure of the sample, by working backwards to generate the original molecule. This technique is frequently used by chemists to identify the structure of an unknown chemical compound. However, serum contains a mixture of many proteins, with widely varying concentrations. In this case, instead of a series of a few sharp peaks, the resulting mass spectrum contains many peaks, of varying height. While one cannot use these data to work backwards and reconstruct the structure of each protein, it can be used to identify patterns of proteins that differ between healthy and diseased patients. Recently a new blood test based on this technique to screen for ovarian cancer received widespread media attention. The test was first described in the medical literature in 2002 [83]. In this test, a blood sample is obtained from a patient. Serum proteins are isolated and the sample is analyzed using mass spectrometry. Scientists obtained blood from 50 women known to Figure 10.46. Mass spectrometry of blood serum may help differentiate healthy individuals from those with cancer [83]. Reprinted with permission from Elsevier (The Lancet, 2002, Vol. 359 No. 9306, pp. 572–577).

have ovarian cancer, 50 women known to be normal and 16 women with benign ovarian disease. They ana-

charged fragments, most of which have a unit positive

lyzed the resulting mass spectrometry data to search for protein peaks which differed in these two groups of patients. They examined thousands of proteins and identified a few which appeared to be different in the two groups. Using these differences they were able to define a diagnostic algorithm which correctly identified

charge. These tiny charged fragments are then sprayed out of a nozzle through a magnetic field into a vacuum chamber. The positively charged fragments are accelerated in the vacuum chamber through a strong magnetic field. The time required for each fragment to travel

50 out of 50 patients with ovarian cancer (sensitivity = 100%), and correctly identified 63 out of 66 women as normal (specificity = 95%). You can easily show that, in this setting, the positive predictive value of this test is 94%, significantly higher than what we calculated for

down this chamber is dependent on the ratio of its mass to charge. The mass spectrometer produces a graph that shows distribution of masses in the sample. A computer program is then used to analyze patterns and distinguish blood from patients with cancer and from those without. Figure 10.46 shows a typical mass spectrograph. The protein fragment mass is indicated on the x-axis,

CA-125 or transvaginal ultrasound.

Do you think it is fair to compare the PPV of this test in this setting to our PPV calculations for CA-125 and transvaginal ultrasound? Why or why not?


Biomedical Engineering for Global Health

Figure 10.47. The initial phase in developing the blood test to screen for ovarian cancer (left) and the subsequent phase, validating the pattern found in the first phase (right). Reprinted with permission from Elsevier (The Lancet, 2002, Vol. 359 No. 9306, pp. 572–577).

Let’s examine the development of this test in more detail. In the initial phase of the study, called pattern discovery (Figure 10.47, left), blood samples were obtained from patients known to have cancer and patients known to be normal. Protein mass spectra were obtained from each of these samples, and investigators examined the spectra. Each one contained the strength of the signal at 15,200 different mass/charge ratios. Different types of data analysis were applied to identify a small group of 5–20 key proteins which differed between the two groups of patients. The proteins were characterized by their mass/charge ratio and their relative abundance. This phase of the study is sometimes called the training phase, because it focuses on narrowing down a large number of data points, to identify a small group which provide diagnostically useful information. However, one limitation of this approach is that the number of proteins measured usually greatly exceeds the numbers of patients participating in the trial. Under these conditions, it is possible that differences in protein abundance between patients with cancer and patients without cancer arise due to simple chance fluctuations, and have nothing to do with the disease process at all [83]. From the thousands of peaks measured, the abundance of protein at only five different mass to charge ratios was found to vary between patients with and without cancer. Figure 10.48 shows data from four patients – two with cancer and two without. There is a very different peak intensity at a mass to charge ratio of 2111 in the spectra from cancer patients compared

Figure 10.48. Spectra illustrating that a difference in mass to charge ratio may differentiate patients with ovarian cancer from those without it [83]. Reprinted with permission from Elsevier (The Lancet, 2002, Vol. 359 No. 9306, pp. 572–577).

to the spectra from unaffected individuals. To guard against the possibility that these fluctuations are due to chance, most clinical trials to test new diagnostic tests use a training phase to optimize the algorithm. Then, a second group of patients is recruited and the diagnostic algorithm is applied to data collected from this group; a phase generally referred to as validation (Figure 10.47, right). The performance of the diagnostic algorithm when applied to data in this validation group

Technologies for early detection and prevention of cancer


Testing: Ovarian Cancer The new test is the first diagnostic tool to emerge from the fledgling field of proteomics, the study of proteins and what they can tell us about a person’s

Figure 10.49. A heat map representation of all the ovarian cancer screening study specimens; the shifted spectra at the bottom raise questions about the experimental protocol of the study [101].

gives the best estimate of how well the algorithm performs [83]. Accounts describing the exciting promise of this new diagnostic test were widely reported in the media (see short article from the February 18, 2002 issue of Newsweek above). At the time, the lead author on the study, Lance Liotta, said, “The most important next goal is validating the promise of these results in large, multi-institutional trials [98].” While the general media responded with enthusiasm to the possibility of a new test which could improve the early detection of cancer, response from the scientific community was much more skeptical. Dr. Eleftherios P. Diamandis, head of clinical biochemistry at Mount Sinai Hospital in Toronto, expressed the concern that, “If you don’t know what you’re measuring, it’s a dangerous black-box technology . . . They are rushing into something and it could be a disaster” [99]. Dr. Nicole Urban, head of gynecologic cancer research at the Fred Hutchinson Cancer Research Center in Seattle warned patients, “Certainly there’s no published work that would make me tell a woman she should get this test” [99]. The datasets used to generate the ovarian cancer screening algorithm were made publicly available. When others tried to reproduce the results reported in the literature, several problems were identified. Most importantly, it appeared that there was a change in the experimental protocol for the measurements made from benign specimens that caused a systematic change in the data. Figure 10.49 shows a heat map representation of the 216 spectra from the pattern discovery and validation phases of the data. The m/z ratio runs along the x-axis, and the samples are grouped by diagnosis. There is a clear difference in the pattern of the benign speci-

disease state. The goal is to identify the protein patterns in the blood that are unique to a disease. Once the researchers had linked ovarian cancer to a characteristic pattern, they tested the method’s predictive power by screening blood samples from 116 women – 50 with cancer and 66 with benign conditions. The new test enabled doctors to pick out all 50 malignancies, including each of the 18 early-stage cases. More trials on larger groups of women are still needed in order to obtain FDA approval. If the test passes muster, it could be available in five years or less. Since it takes only 30 minutes to perform, it should be easily affordable. But the real savings could be in lives: 14,000 that are lost every year to this deadly disease. Anne Underwood. From Newsweek, February 18  c 2002. Newsweak Inc. All rights reserved. Used with permission and protected by the Copyright laws of the United States. The printing, copying, redistribution, or retransmission of the Material without express written permission is prohibited.

mens shown at the bottom of the figure possibly due to a change in protocol for these specimens [101]. The use of proteomics technology at present can be thought of as a “black box technology.” Serum samples are sent into the black box, and a diagnosis comes out. Because the approach does not rely on biological explanations, it is crucial that the approach be reliable and reproducible in any location. Further studies with additional samples are required to demonstrate the potential of this new technology.

Summary In this chapter, we have seen the benefits and possible harms associated with cancer screening. Screening


Biomedical Engineering for Global Health

should be undertaken only when the following conditions have been met: (1) the effectiveness of the screening test has been demonstrated, (2) there are sufficient economic resources to screen all patients in the target group, (3) there are tools to confirm disease in patients with a positive screening test, (4) there are existing procedures to treat the disease, (5) and when disease prevalence is high enough to justify effort and costs of screening. One of the challenges of screening is that it may reveal disease that might never be detected or cause problems otherwise. This is certainly true for screening for cervical cancer with the Pap test. Most abnormalities found on the Pap smear never become invasive cancer. However, there are relatively low cost, minimally invasive tools to follow an abnormal Pap smear, and treatment of high grade pre-cancer can prevent future development of cervical cancer. The use of screening has dramatically reduced both the incidence and the mortality of cervical cancer throughout the developed world. Likewise, screening for prostate cancer likely identifies many cases of prostate cancer which would otherwise have never produced any symptoms. Unlike the case of cervical cancer, screening with the PSA test has dramatically increased the apparent incidence of prostate cancer, while the mortality has largely remained unchanged. Ovarian cancer presents one of the most difficult challenges in cancer screening; because it is a relatively rare disease, any potential screening test must have a very high specificity to yield a reasonable predictive value. The relative inaccessibility of the ovaries makes it difficult and invasive to follow up an abnormal screening test. As a result, we do not currently screen for ovarian cancer although it is the most deadly of the female reproductive cancers. While screening can have important medical benefits, it requires resources, both to test people and to follow up abnormal screening results. How do we decide if screening represents a good investment of healthcare resources? In the next chapter, we will examine the use of cost effectiveness analysis to make these decisions.

Bioengineering and Global Health Project Project task 6: Gather information regarding current research and development efforts What research and development efforts are currently underway to solve the health problem that you have identified? Write a one-page summary of this research, summarizing what is known about the effectiveness or limitations of these current procedures.

Homework 1. In the USA, what is the most prevalent cancer in (a) men and (b) women? Worldwide, what is the most prevalent cancer in (c) men and (d) women? 2. Cancer screening. a. What four types of cancer are routinely screened for in the United States? For each, describe the screening test that is used. b. Do most people in the USA adhere to screening recommendations? What factors cause people not to be screened? c. Discuss whether these screening tests are used throughout the rest of the world. 3. Lung cancer is the leading cause of cancer death for both men and women in the United States. More people die as a result of lung cancer than of colon, breast, and prostate cancers combined. Lung cancer is rare in people under the age of 40. The average age of people diagnosed with lung cancer is 60. In 2004 there are expected to be about 173,770 new cases of lung cancer in the United States [102]. About 160,440 people will die of this disease. The population of the United States in 2004 is 292,287,454. a. Calculate the annual incidence rate of lung cancer in the USA in 2004. b. Calculate the mortality rate of lung cancer in the USA in 2004. c. Why is the mortality rate of lung cancer so high? 4. Describe in your own words, WITHOUT using equations or other mathematical expressions or the

Technologies for early detection and prevention of cancer words “true”, “false”, “positive”, or “negative” the following terms with regard to a screening test for ovarian cancer. True Positive False Positive False Negative True Negative PPV NPV 5. A diagnostic test is 92% sensitive and 94% specific. A test group is comprised of 500 people known to have the disease and 500 people known to be free of the disease. How many of the known positives would actually test positive? How many of the known negatives would actually test negative? 6. A screening test for a particular disease has a


information, and they cite disease population screening data in the literature which reports that 6% of that population was positive when screened. Referring to the literature, you discover that the screening test used had sensitivity of 95% and specificity of 98%. What proportion of the population over 50 years of age do you think really has the disease? Source: [103]. 11. A recent study examined the expression of p53 (a protein found in many transformed cell lines derived from tumors) as a marker for ovarian cancer. The sensitivity and specificity of p53 as a marker for the diagnosis of ovarian cancer in this study were 82% and 93% respectively. Forty-seven patients with no family history of breast or ovarian cancer were included in the study. Fourteen of the 17 patients with ovarian

sensitivity of 96% and a specificity of 92%. You plan to screen a population in which the prevalence of the disease is 0.2%. How many false positives will be found by this screening procedure

cancer had p53 overexpression. Fifteen of the 47 patients had never given birth. a. If p53 overexpression was used as a test for ovarian cancer, how many patients in this study

for each true positive that is found? 7. A clinical trial of a new automated mammography system was carried out in 50,000 women known to

received a false positive test result? b. If p53 overexpression was used as a test for ovarian cancer, how many patients in this study

have breast cancer. If 37,500 women received a positive test result, what would the specificity of the new test be? 8. Based on all the information currently available, you estimate that the patient in your office has a one in four chance of having a serious disease. You order a diagnostic test with sensitivity of 95% and a specificity of 90%. The result comes back positive. Based on all the information available, what would be the chance your patient really has the disease? Source: [103]. 9. A test with 99.9% sensitivity and 99% specificity is used to screen a population for a disease with 1% prevalence. What would be the proportion of test positives in the screen who actually have the disease? Source: [103]. 10. The American Disease X Foundation reports that 6% of the population over 50 years of age has Disease X. You inquire as to the source of their

received a false negative test result? c. How much better are these results for a screening test than CA-125? 12. You are a physician for Mr. Jones, a 65 year old African American man who presents to you with complaints of difficulty urinating. Specifically, he has trouble starting urine flow and has an intermittent stream. He says he noticed this problem some time ago, and that it has slowly been getting worse. Mr. Jones says he has always been healthy and has not seen a doctor in thirty years. He was adopted and does not know his family history. a. What disease discussed in Chapter 10 might explain Mr. Jones’ symptoms? b. What three risk factors for this disease does Mr. Jones have? c. What initial tests are available that might aid your diagnosis of Mr. Jones?


Biomedical Engineering for Global Health d. If the initial tests are positive, what would be the next step in diagnosis? e. Mr. Jones does indeed have the disease you suspected, and you recommend surgical intervention. Any surgical procedure has the risks of pain, bleeding, and infection. What are two specific risks associated with this particular surgery? f. List two reasons why the tests listed in part c are controversial for use as screening tools. g. In American males, prostate cancer is the most common, non-skin cancer (accounting for 33% of all new cancers), but is less deadly than might be expected, ranking behind both lung and colon cancer as the third leading cause of cancer death (9%). By contrast, ovarian cancer is the eighth most common new non-skin cancer in American women (3% of new diagnoses), but accounts for a surprising number of deaths; it ranks as the fifth leading cause of cancer death in this population (6%). Give three reasons for the discrepancy between the incidence and death rates for these two

cancer types. 13. A patient comes to your office complaining of abdominal fullness and a change in bowel habits. She reports a family history of breast cancer and ovarian cancer. You suspect she may have ovarian cancer and order a serum CA-125 test. The sensitivity of this test is 35% and the specificity is 98.5%. The incidence of ovarian cancer in this population is 0.1%. The test comes back positive. a. If you gave this test to 1,000,000 women, how many patients would have a true positive (TP) result, a false positive (FP) result, a true negative (TN) result and a false negative (FN) result? b. Given her positive test result, what is the likelihood that your patient really has ovarian cancer? c. What test would you recommend that your patient undergo next?

14. A company called BioCurex recently announced results of a clinical trial for a new test to detect lung cancer (see story below). RANCHO SANTA MARGARITA, Calif.–(BUSINESS WIRE)–April 5, 2004–BioCurex Inc. announces results for lung cancer detection using its proprietary Serum-RECAF(TM) blood test. The results confirm 90% sensitivity with 95% specificity. The findings further substantiate the use of RECAF(TM) as a universal cancer marker with a potential market size of $2 billion per year for all cancers. The study included 32 lung cancer patients and 103 normal donors with statistical verification.  c

2003 Business Wire. [ news_story.cfm?StoryID=15650520&full=1] a. Calculate the number of patients with true negative (TN), true positive (TP), false positive (FP) and false negative (FN) test results in this trial. b. What is the positive predictive value in this trial? c. Do you think the PPV you calculated in part b is an accurate estimate of what to expect if the test is used to screen the general population for lung cancer? Why or why not? 15. Suppose we have two new screening tests for ovarian cancer – Test A and Test B. When tested in a large population, we find the sensitivity and specificity values for the two tests listed in the table below. Your mother is worried about her risk of ovarian cancer because both her mother and sister died of ovarian cancer at a young age. She asks your advice about which screening test to undergo. Which test would you recommend that she take? Why?




Test A



Test B



Technologies for early detection and prevention of cancer 16. Consider the development of a new proteomics based screening test for ovarian cancer described in this chapter. Apply the five steps of technology assessment to this new technology. Does this assessment support the use of the technology?

References [1] American Cancer Society. Cancer Facts and Figures 2007. Atlanta: American Cancer Society; 2007. ˜ AM, Heron MP, Smith BL. Deaths: Preliminary [2] Minino Data for 2004. National Vital Statistics Report. 2006 June 28;54(19). [3] American Cancer Society. Cancer Facts and Figures 2005. Atlanta: American Cancer Society; 2005. [4] Edwards BK, Brown ML, Wingo PA, Howe HL, Ward E, Ries LAG, et al. Annual Report to the Nation on the Status of Cancer, 1975–2002, Featuring PopulationBased Trends in Cancer Treatment. Journal of the National Cancer Institute. 2005 October 5; 97(19). [5] WHO. Mortality: Revised Global Burden of Disease (2002) Estimates. Geneva: World Health Organization; 2002. [6] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA: a Cancer Journal for Clinicians. 2005 Mar–Apr; 55(2): 74–108. [7] International Union Against Cancer. Global Action Against Cancer. Geneva: WHO/UICC; 2005. [8] World Health Organization. Global Cancer Rates Could Increase by 50% to 15 Million by 2020. Geneva: World Health Organization; 2003 April 3. [9] American Cancer Society. The Case for Global Action. 2007 [cited 2007 May 9]; Available from: AA_2_5_1x_Case_for_Global_Action.asp?sitearea=AA [10] American Cancer Society. Lifetime Probability of Developing or Dying from Cancer. 2006 February 22 [cited 2007 May 9]; Available from: 6x_Lifetime_Probability_of_Developing_or_Dying_ From_Cancer.asp?sitearea= [11] Stewart BW, Kleihues P, eds. World Cancer Report. Lyon: IARC Press; 2003. [12] American Cancer Society. Cancer Statistics 2005 [PowerPoint]. Atlanta: American Cancer Society; 2005. [13] MacKay J, Eriksen M. The Tobacco Atlas. Geneva: The World Health Organization; 2002. [14] American Cancer Society. Cancer Prevention and Early Detection Facts and Figures 2007. Atlanta: American Cancer Society; 2007.


[15] Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000 Jan 7; 100(1): 57–70. [16] Dolan DNA Learning Center – Cold Spring Harbor Laboratory. Inside Cancer. [Multimedia Guide] 2004 [cited 2007 May 10]; Available from: [17] Cotran RS, Kumar V, Robbins SL. Robbins Pathologic Basis of Disease. 5th edn. W.B. Saunders Company; 1994. [18] National Cancer Institute. Milestone (1971): President Nixon declares war on cancer. 2007 [cited 2007 May 10]; Available from: M4_Nixon.htm [19] Niederhuber JE. The Nation’s Investment in Cancer Research. Frederick, Maryland: National Cancer Institute; 2006 October. [20] National Cancer Institute. The NCI Strategic Plan for Leading the Nation. Frederick, Maryland: National Cancer Institute; 2006 January. [21] National Center for Health Statistics. Health, United States, 2006 With Chartbook on Trends in the Health of Americans. Hyattsville, MD; 2006. [22] Carr DB, Loeser JD, Morris DB. Narrative, Pain and Suffering (Progress in Pain Research and Management). Seattle, WA: IASP Press; 2005. [23] Wang XS, Li TD, Yu SY, Gu WP, Xu GW. China: status of pain and palliative care. Journal of Pain and Symptom Management. 2002 Aug; 24(2): 177–9. [24] US Preventive Services Task Force. Guide to Clinical Preventive Services. 2nd edn. Alexandria, VA: International Medical Publishing; 1996. [25] Elmore JG, Armstrong K, Lehman CD, Fletcher SW. Screening for breast cancer. Jama. 2005 Mar 9; 293(10): 1245–56. [26] Warner E, Plewes DB, Hill KA, Causer PA, Zubovits JT, Jong RA, et al. Surveillance of BRCA1 and BRCA2 mutation carriers with magnetic resonance imaging, ultrasound, mammography, and clinical breast examination. Jama. 2004 Sep 15; 292(11): 1317–25. [27] Thomas DB, Gao DL, Ray RM, Wang WW, Allison CJ, Chen FL, et al. Randomized trial of breast self-examination in Shanghai: final results. Journal of the National Cancer Institute. 2002 Oct 2; 94(19): 1445–57. [28] American Cancer Society. Cancer Facts and Figures 2006. Atlanta: American Cancer Society; 2006. [29] Humphrey LL, Helfand M, Chan BK, Woolf SH. Breast cancer screening: a summary of the evidence for the







[35] [36]






Biomedical Engineering for Global Health US Preventive Services Task Force. Annals of Internal Medicine. 2002 Sep 3; 137(5 Part 1): 347–60. Parkin DM, Bray F. Chapter 2: The burden of HPVrelated cancers. Vaccine. 2006 Aug 21; 24 Suppl. 3: S11–25. Nijhuis ER, Reesink-Peters N, Wisman GB, Nijman HW, van Zanden J, Volders H, et al. An overview of innovative techniques to improve cervical cancer screening. Cell Oncology. 2006; 28(5–6): 233–46. Burchell AN, Richardson H, Mahmud SM, Trottier H, Tellier PP, Hanley J, et al. Modeling the sexual transmissibility of human papillomavirus infection using stochastic computer simulation and empirical data from a cohort study of young women in Montreal, Canada. American Journal of Epidemiology. 2006 Mar 15; 163(6): 534–43. American Cancer Society. FDA Approves New Cervical Cancer Screening Test. 2003 March 31 [cited 2007 May 10]; Available from: NWS/content/NWS_1_1x_FDA_Approves_New_ Cervical_Cancer_Screening_Test.asp Cohen J. Public health. High hopes and dilemmas for a cervical cancer vaccine. Science (New York, NY). 2005 Apr 29; 308(5722): 618–21. Alberts B. Molecular Biology of the Cell. 3rd edn. New York: Garland Publishing; 1994. Renshaw AA. Measuring sensitivity in gynecologic cytology: a review. Cancer. 2002 Aug 25; 96(4): 210–17. Centers for Disease Control. Regulations for Implementing the Clinical Laboratory Improvement Amendments of 1988: a Summary. MMWR. 1992 (No. RR-2). Austin RM. Human papillomavirus reporting: minimizing patient and laboratory risk. Archives of Pathology & Laboratory Medicine. 2003 Aug; 127(8): 973–7. Bogdanich W. Lax Laboratories: The Pap Test misses much cervical cancer through lab’s errors. The Wall Street Journal. 1987 November 2. Nanda K, McCrory DC, Myers ER, Bastian LA, Hasselblad V, Hickey JD, et al. Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic abnormalities: a systematic review. Annals of Internal Medicine. 2000 May 16; 132(10): 810–19. Fahey MT, Irwig L, Macaskill P. Meta-analysis of Pap test accuracy. American Journal of Epidemiology. 1995 Apr 1; 141(7): 680–9.

[42] Mitchell MF, Schottenfeld D, Tortolero-Luna G, Cantor SB, Richards-Kortum R. Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis. Obstetrics and Gynecology. 1998 Apr; 91(4): 626–31. [43] Welch HG. Should I be Tested for Cancer? Berkely, CA: University of California Press; 2004. [44] Ferenczy A, Franco E. Cervical-cancer screening beyond the year 2000. The Lancet Oncology. 2001 Jan; 2(1): 27–32. [45] Hutchinson ML. Assessing the costs and benefits of alternative rescreening strategies. Acta Cytologica. 1996 Jan–Feb; 40(1): 4–8. [46] Kurman RJ, Henson DE, Herbst AL, Noller KL, Schiffman MH. Interim guidelines for management of abnormal cervical cytology. The 1992 National Cancer Institute Workshop. Jama. 1994 Jun 15; 271(23): 1866–9. [47] Jeronimo J, Morales O, Horna J, Pariona J, Manrique J, Rubinos J, et al. Visual inspection with acetic acid for cervical cancer screening outside of low-resource settings. Revista Panamericana de Salud Publica (Pan American Journal of Public Health). 2005 Jan; 17(1): 1–5. [48] Birdsong GG. Automated screening of cervical cytology specimens. Human Pathology. 1996 May; 27(5): 468–81. [49] Groopman J. Contagion: Papilloma Virus. The New Yorker. 1999 September 13. [50] Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, et al. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA: a Cancer Journal for Clinicians. 2002 Nov–Dec; 52(6): 342–62. [51] Russell J, Crothers BA, Kaplan KJ, Zahn CM. Current cervical screening technology considerations: liquid-based cytology and automated screening. Clinical Obstetrics and Gynecology. 2005 Mar; 48(1): 108–19. [52] Health Technologies Advisory Committee – Minnesota. New technologies for cervical cancer screening: strategies to lower the rate of cervical cancer. Health Services/Technology Assessment Text (HSTAT). National Library of Medicine; 1999. [53] FDA. FDA Approves Expanded Use of HPV Test. 2003 March 31 [cited 2007 May 20]; Available from: NEW00890.html [54] Schiffman M, Herrero R, Hildesheim A, Sherman ME, Bratti M, Wacholder S, et al. HPV DNA testing in

Technologies for early detection and prevention of cancer












cervical cancer screening: results from women in a high-risk province of Costa Rica. Jama. 2000 Jan 5; 283(1): 87–93. Arbyn M, Sasieni P, Meijer CJ, Clavel C, Koliopoulos G, Dillner J. Chapter 9: Clinical applications of HPV testing: a summary of meta-analyses. Vaccine. 2006 Aug 21; 24 Suppl. 3: S78–89. R 2 High-Risk HPV Digene Corporation. Hybrid Capture DNA TestTM [cited 2007 May 20]; Product Insert]. Available from:,%20hc2%20HPV%20DNA%20Test%20US.pdf Germar MJ, Marialdi M. Visual inspection with acetic acid as a cervical cancer screening tool for developing countries. Review for the 12th Postgraduate Training Course in Reproductive Health/Chronic Disease 2003 [cited 2007 May 28]; Available from: http://www. Visual_inspection_cervical_ cancer.htm Coalition NCC. NCCC Section on the HPV Vaccine. 2007 [cited 2007 March 2]; Available from: Cohen J. High hopes and dilemmas for a cervical cancer vaccine. Science (New York, NY). 2003; 308: 618–21. Kane MA, Sherris J, Coursaget P, Aguado T, Cutts F. Chapter 15: HPV vaccine use in the developing world. Vaccine. 2006 Aug 21; 24 Suppl. 3: S132–9. Sankaranarayanan R, Shastri SS, Basu P, Mahe C, Mandal R, Amin G, et al. The role of low-level magnification in visual inspection with acetic acid for the early detection of cervical neoplasia. Cancer Detection and Prevention. 2004; 28(5): 345–51. Pan American Health Organization. Visual Inspection of the Uterine Cervix with Acetic Acid (VIA): A Critical Review and Selected Articles. Washington, D.C.; 2003. Park SY, Follen M, Milbourne A, Rhodes H, Malpica A, MacKinnon M, et al. Automated image analysis of digital colposcopy for the detection of cervical neoplasia. Manuscript submitted for publication, 2007. Kasper D, Braunwald E, Fauci A, Longo D, Hauser S, Jameson JL, eds. Harrison’s Principles of Internal Medicine. 16th edn. New York City: McGraw-Hill; 2005. Hutchinson K, Wener M. National Health and Nutrition Examination Survey Lab Methods 2003–2004: Total Prostate-Specific Antigen (PSA). University of Washington Medical Center; 2007 May 4.


[66] Aureon Biosystems. LumiPhos Plus. [cited 2007 March 10]; Available from: [67] Lumigen. Lumigen PPD. 2006 [cited 2007 March 10]; Available from: LumigenPPD.shtml [68] Hernandez J, Thompson IM. Prostate-specific antigen: a review of the validation of the most commonly used cancer biomarker. Cancer. 2004 Sep 1; 101(5): 894–904. [69] Albertsen PC. Is screening for prostate cancer with prostate specific antigen an appropriate public health measure? Acta Oncologica (Stockholm, Sweden). 2005; 44(3): 255–64. [70] Gottlieb RH, Mooney C, Mushlin AI, Rubens DJ, Fultz PJ. The prostate: decreasing cost-effectiveness of biopsy with advancing age. Investigative Radiology. 1996 Feb; 31(2): 84–90. [71] American Cancer Society. How Is Prostate Cancer Diagnosed? 2006 July 26 [cited 2007 May 20]; Available from: CRI_2_4_3X_How_is_prostate_cancer_diagnosed_36.asp [72] Lu-Yao GL, Yao SL. Population-based study of long-term survival in patients with clinically localised prostate cancer. The Lancet. 1997 Mar 29; 349(9056): 906–10. [73] Neal DE, Leung HY, Powell PH, Hamdy FC, Donovan JL. Unanswered questions in screening for prostate cancer. European Journal of Cancer. 2000 Jun; 36(10): 1316–21. [74] Bartsch G, Horninger W, Klocker H, Reissigl A, Oberaigner W, Schonitzer D, et al. Prostate cancer mortality after introduction of prostate-specific antigen mass screening in the Federal State of Tyrol, Austria. Urology. 2001 Sep; 58(3): 417–24. [75] Horninger W, Berger A, Pelzer A, Klocker H, Oberaigner W, Schonitzer D, et al. Screening for prostate cancer: updated experience from the Tyrol study. The Canadian Journal of Urology. 2005 Feb; 12 Suppl. 1: 7–13; discussion 92–3. [76] Ilic D, O’Connor D, Green S, Wilt T. Screening for prostate cancer: a Cochrane systematic review. Cancer Causes Control. 2007 Apr; 18(3): 279–85. [77] Amling CL. Prostate-specific antigen and detection of prostate cancer: what have we learned and what should we recommend for screening? Current Treatment Options in Oncology. 2006 Sep; 7(5): 337–45.


Biomedical Engineering for Global Health

[78] Hennekens CH, Buring JE. Epidemiology In Medicine. Philadelphia: Lippincott Williams & Wilkins; 1987. [79] Welch HG, Schwartz LM, Woloshin S. Are increasing 5-year survival rates evidence of success against cancer? Jama. 2000 Jun 14; 283(22): 2975–8. [80] Reynolds EA, Moller KA. A review and an update on the screening of epithelial ovarian cancer. Current Problems in Cancer. 2006 Sep–Oct; 30(5): 203–32. [81] Rosenthal AN, Menon U, Jacobs IJ. Screening for ovarian cancer. Clinical Obstetrics and Gynecology. 2006 Sep; 49(3): 433–47. [82] BC Cancer Agency. Years of Life Lost. 2005 July 25 [cited 2007 May 22]; Available from: LifeLost.htm [83] Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA, Steinberg SM, et al. Use of proteomic patterns in serum to identify ovarian cancer. The Lancet. 2002 Feb 16; 359(9306): 572–7. [84] Urban N. Specific keynote: ovarian cancer risk assessment and the potential for early detection. Gynecologic Oncology. 2003 Jan; 88(1 Pt 2): S75–9; discussion S80–3. [85] Nahhas WA. Ovarian cancer. Current outlook on this deadly disease. Postgraduate Medicine. 1997 Sep; 102(3): 112–20. [86] Zurawski VR, Jr., Sjovall K, Schoenfeld DA, Broderick SF, Hall P, Bast RC, Jr., et al. Prospective evaluation of serum CA 125 levels in a normal population, phase I: the specificities of single and serial determinations in testing for ovarian cancer. Gynecologic Oncology. 1990 Mar; 36(3): 299–305. [87] Gladstone CQ. Screening for ovarian cancer. In: Canadian Task Force on Periodic Health Examination, ed. Canadian Guide to Clinical Preventive Health Care. Ottawa: Health Canada; 1994: 870–81. [88] Menon U, Jacobs IJ. Ovarian cancer screening in the general population. Current Opinion in Obstetrics & Gynecology. 2001 Feb; 13(1): 61–4. [89] van Nagell JR, Jr., DePriest PD, Reedy MB, Gallion HH, Ueland FR, Pavlik EJ, et al. The efficacy of transvaginal sonographic screening in asymptomatic women at risk for ovarian cancer. Gynecologic Oncology. 2000 Jun; 77(3): 350–6. [90] Lok IH, Sahota DS, Rogers MS, Yuen PM. Complications of laparoscopic surgery for benign ovarian cysts. The Journal of the American Association









[99] [100] [101]



of Gynecologic Laparoscopists. 2000 Nov; 7(4): 529–34. Hensley ML, Spriggs DR. Cancer screening: how good is good enough? Journal of Clinical Oncology. 2004 Oct 15; 22(20): 4037–9. Johnson D, Sandmire D, Klein D. Ovarian Cancer. In Medical Tests That Can Save Your Life: 21 Tests Your Doctor Won’t Order Unless You Know to Ask. Emmaus, PA Rodale; 2004. Fees Involved in Infertility Treatments. 2005 [cited 2007 June 1]; Available from: Jacobs IJ, Skates S, Davies AP, Woolas RP, Jeyerajah A, Weidemann P, et al. Risk of diagnosis of ovarian cancer after raised serum CA 125 concentration: a prospective cohort study. BMJ (Clinical Research Ed.) 1996 Nov 30; 313(7069): 1355–8. Jacobs IJ, Skates SJ, MacDonald N, Menon U, Rosenthal AN, Davies AP, et al. Screening for ovarian cancer: a pilot randomised controlled trial. The Lancet. 1999 Apr 10; 353(9160): 1207–10. UK Collaborative Trial of Ovarian Cancer Screening. 2006 July 11 [cited 2007 May 20]; Available from: Buys SS, Partridge E, Greene MH, Prorok PC, Reding D, Riley TL, et al. Ovarian cancer screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial: findings from the initial screen of a randomized trial. American Journal of Obstetrics and Gynecology. 2005 Nov; 193(5): 1630–9. FDA. Protein Patterns May Identify Ovarian Cancer. 2002 February 7 [cited 2007 May 20]; Available from: NEW00797.html Pollack A. New cancer test stirs hope and concern. The New York Times. 2004 February 3. Underwood A. Testing: ovarian cancer. Newsweek. 2002 February 18: 12. Baggerly KA, Morris JS, Coombes KR. Reproducibility of SELDI-TOF protein patterns in serum: comparing datasets from different experiments. Bioinformatics (Oxford, England). 2004 Mar 22; 20(5): 777–85. American Cancer Society, How Many People Get Lung Cancer, 2004; Available from: docroot/CRI/content/CRI_2_2_IX_How_many_people_ get_lung_cancer_26.asp?sitearea= ReviewQuestions/SensitivitySpecificity_Questions.htm.

11 Cost effectiveness of screening for disease

We began this book by considering the case study of a new treatment for advanced breast cancer – high dose chemotherapy and bone marrow transplant. Motivated by the disappointing performance of this treatment when it was offered prematurely to thousands of American women, we considered how to systematically evaluate new technologies in Chapter 2. Recall that our process of technology assessment consists of asking five questions about a new medical technology [1]. Biologic plausibility. Does our current understanding of the biology of the disease in question support the use of the technology? Technical feasibility. Can we safely and reliably deliver the new technology to the target patients? Clinical trials. Do the results of randomized clinical trials comparing the new technology to current standards of care show a benefit? Patient outcomes. Are patients better off for having used the new technology? Societal outcomes. What are the costs and ethical implications of the technology?

in the development of new technologies. However, we have not yet developed the tools to determine whether a new technology is a cost effective use of healthcare resources. In Chapter 10, we considered the pros and cons of screening for early cancer. We found that while screening for cervical cancer can reduce both cancer incidence and mortality, there is considerable debate about whether to screen for prostate cancer. In both cases, we must make decisions about how to use limited healthcare resources. Potentially, screening can identify disease at an earlier stage, when it is less expensive to treat. If effective screening tests are available, we must decide who should be screened, how frequently screening should occur, and what test (or test combination) provides the most effective use of resources. In the present chapter, we will learn how to use cost effectiveness analysis to help make recommendations about whether new technologies should be adopted and how they can be used most effectively.

Types of economic evaluation of health technology

Thus far, we have examined the biological plausibility and technical feasibility of new technologies to prevent

How do we approach the economic evaluation of a new health technology? In general, we are interested in com-

infectious disease and to detect cancers at an early stage. We have seen the crucial role that clinical trials play and the importance of adhering to ethical guidelines

paring both the benefits and costs of a new health technology relative to some standard of care. In order to make this comparison meaningful, we must calculate


Biomedical Engineering for Global Health (e.g., blood pressure, number of strokes, cases of cancer found, life years gained, etc.). Cost effectiveness ratios can be computed and compared, and may be expressed for example as dollars spent per case of cancer detected, or dollars spent per life year gained. The advantage of this approach is that it is generally simpler to calculate this type of outcome. The challenge of this approach is that it is difficult to compare the cost effectiveness of two very different types of clinical strategies (e.g. vaccination vs. end stage cancer therapy). An alternative approach is to evaluate clinical benefits in terms of a common set of units. Quality adjusted

Figure 11.1. A representation of health-policy space, comparing the costs and health benefits of interventions.

life years are usually considered as the “yardstick” for this approach, since they are comparable across inter-

the costs and benefits of the technologies in a way that can be compared. Figure 11.1 shows one way of think-

ventions. Recall that the quality adjusted life year is the number of years of full health that are considered equivalent to a given number of years in a reduced

ing about this comparison. This figure represents what is sometimes called “health-policy space.” One axis of the graph represents health, while the other represents

state of health. The advantage of this approach is that expressing cost effectiveness in a common set of units is extremely useful for making policy planning decisions.

cost. All health interventions fall in one of the four quadrants of this graph. In the lower right quadrant, we have interventions which both cost money and worsen health. Obviously we do not want to adopt such inter-

The disadvantage is that they are more difficult to calculate than physical units.

ventions. Technologies which save money but worsen health fall in the lower left quadrant; frequently there is difficult public debate associated with whether to adopt new interventions which may slightly worsen health but have potential for enormous cost savings. In the upper left quadrant, we have interventions which both save money and improve health. We obviously want to adopt interventions which fall in this quadrant; unfortunately, there are not many interventions which fall here. Many childhood vaccines are one of the few interventions which both improve health and save money. Most interventions fall in the top right quadrant, improving health but consuming resources. How do we determine which quadrant of the graph a new intervention falls in? And how can we compare two interventions within a quadrant? There are several approaches to evaluate the cost effectiveness of an intervention; here we will consider two in detail. The first method is to compare clinical strategies in terms of clinical outcomes as measured in physical units

Components of an economic evaluation for health technology assessment Just like the method of engineering design, we can follow a step-wise process to carry out an economic evaluation of a health technology. We first consider the steps in this process, and then turn to several examples. You will see that the economic evaluation includes many components in addition to the “economics”; in other words, to appropriately perform an economic evaluation, one must also perform an evaluation of the effectiveness of a technology as well.

Define the problem This is the starting point for an economic evaluation. A problem should be well defined and posed in an answerable form [2]. Typically a trade-off is usually obvious and should be made explicit, e.g. if improved survival rates have been observed for an expensive intervention, one would like to determine if the improvement in clinical outcomes is worth the additional cost from the intervention. Similarly, if there are trade-offs between quality

Cost effectiveness of screening for disease


Healthcare outside the clinic: July 5, 2007 Tessa Swaziland After two days of WFP work, I was able to escape the clinic and see a bit more of the healthcare facilities outside of the Baylor world. Wednesday, I went with Carrie and Julia to Vuvulane, a rural community about two hours from the COE. The clinic was run by nurses, and the only doctors ever out there were the Baylor doctors once a month. We met up with Good Shepherd (hospital in Manzini) nurses who have been going to these outreach sites before Baylor got involved. They provided the connections and the ARVs. We provided the medical expertise. I spent the day with one of the nurses. I was “helping” with pill-counting, but really I was just watching. The nurse was super nice and seemed to think I was just great, so we had quite a bit of fun while we worked. As I watched her work, I kept noticing things she was doing incorrectly (as she filled out the adherence sheet). The whole point of pill-counting is to see how well a patient is adhering to their drug regimen. It is very important because if a patient has poor adherence, then their virus will build resistance to the drug, and the patient’s viral load will increase. (Basically, they will get sicker, and they will be much harder to treat in the future.) So, I started asking her questions about why and how she did what she did. Like, when she wrote down 56 for the expected number of pills, I asked why. She said that the patient needed two a day, and since there are 28 days between visits, they should’ve taken 56. Looking at the records, I could see that there were 30 days between this visit and the last. When I asked her about that, she said, “We always do 28. That’s just what we do.” And that was that. There were quite a few other major problems I observed with the system, but ultimately, it was all meaningless, since they never calculated their adherence percent. And even if the person clearly had terrible adherence (one man had 50 extra pills!), they never counseled them or took them off the ART (anti-retroviral treatment). Drug resistance has implications for not only the individual who isn’t taking his meds correctly, but also anyone he passes the disease to, who will suffer from drug-resistant HIV. On the way to and from Vuvulane, we drove by Hlane National Park, where Carrie was lucky enough to see a lion. I was rummaging through my bag and missed it completely! On the way back, we stopped at the private Mbabane hospital, which was much nicer than the government hospital but still had much to improve upon before it would ever meet American standards. We were checking up on a premature infant who was born at 28 weeks. (I could be off a few weeks there) I had never seen a preemie before and wasn’t quite prepared for it. The mother even asked me if I was afraid of the baby because I stayed by the door and stayed there throughout the visit. The baby was miniscule – the diaper engulfed its entire body, and its leg was the thickness of my thumb. What I was afraid of was giving the baby some germ from the outside world that would do its premature immune system in. Carrie told me that a neo-natal unit in the states would be 10 times quieter and darker to simulate the conditions in the womb. This poor baby had a bright beam of sunlight pouring over him, in addition to a symphony of construction sounds reverberating from somewhere nearby in the hospital. Julia said that if the baby had been in the government hospital, there is no question that it would be dead by now. On Thursday, I shadowed Dr. Eileen at the government hospital. Although I’ve never been interested in clinical medicine, and I never will be, I found rounds to be quite interesting. She and Dr. D (doctor at that hospital) went from bed to bed examining patients, asking the mothers questions, and looking at X-rays. One of the biggest challenges was deciding whether they had a weakened immune system due to AIDS or TB or both. Sometimes the x ray clearly indicated TB, but with some patients, it was more difficult to tell. One of the most exciting things about our visit was the fact that, when we arrived, Dr. D and some nurses had sat down all of the patient’s mothers (this was the pediatric ward) and were discussing the importance of washing hands and boiling water. This may seem insignificant, but in fact, if they can effectively communicate these messages to the mothers, they could more than halve the number of patients that needed to be there. They plan to do a lot more of this sort of doctor-patient interaction,


Biomedical Engineering for Global Health and they were also trying to make signs to accompany the campaign. In that respect, I was able to help. Making those signs is one of the smaller projects I’ve been working on. On Friday, I presented an updated plan for WFP. Thursday, we had a volunteer come in, and Dave started training her. The plan for the following week was to transition from the old system (where Dave and I are very involved) to the new system (where Dave and I won’t be here to do anything). A lot of the doctors weren’t there, so that kind of caused problems, but nothing too big.

Steps in a cost effectiveness assessment Define the problem Identify the perspective Identify the alternatives Analyze the effectiveness Analyze the costs Perform discounting Perform sensitivity analysis Address ethical issues Interpret the results

and quantity of life, these should be noted explicitly as well.

Identify the perspective In carrying out an economic analysis, one must consider whose costs to include; this is called the perspective of the analysis. The perspective is of critical importance when performing an economic evaluation. The ideal perspective is that of the “societal perspective,” which incorporates all aspects of costs and benefits. In technology assessment, the perspective must clearly be identified, for it is possible to have different results of an analysis when different perspectives are taken. Potential perspectives for an analysis include: patient, health maintenance organization, payer, or society. Guidelines from the Panel on Cost-Effectiveness in Health and Medicine recommend that the “reference case” be done from the societal perspective.

Identify the alternatives Identifying the alternative technologies is a critical component for an economic evaluation of a new technology [3]. Sometimes the baseline strategy is the strategy of doing nothing. Other times, we will use the current standard of care as the baseline strategy. It is also important to note that a “do nothing” strategy may or may not be an appropriate alternative to consider. Doing nothing has certain ethical implications, both from clinical and policy perspectives. “Doing nothing” is likely to have both economic and clinical consequences. In particular, it is rare that a “do nothing” strategy has no economic costs. Certainly, if “doing nothing” means that a patient will die, there are likely to be costs associated with the dying process, such as palliative care. Less dramatically, if “doing nothing” means “no screening”, then associated costs of undetected disease must be incorporated in the analysis.

Analyze the effectiveness The effectiveness is the clinical outcome to be evaluated in an economic evaluation. Ideally, the effectiveness is evaluated in terms of quality adjusted life years. Quality adjusted life years represent a standardized measure of health outcome that incorporates both length and quality of life into consideration. The number of quality adjusted life years is the number of years in perfect health that are valued comparably to the number of life years experienced in a less desirable health state. Similarly, the number of quality adjusted life years is the

Cost effectiveness of screening for disease number of years of life weighted by a measure of the quality of life experienced. The quality of life component is based on a 0–1 utility scale, where 0 represents death and 1 represents perfect health [3, 4].

Analyze the costs Costs can be classified in several categories [5]. Direct healthcare costs include physician services, pharmaceuticals, tests, inpatient care, outpatient care, and administration regarding clinical facilities, including medical records, food, nursing, and supplies. Non-healthcare costs are costs that are borne as a result of seeking medical care. These may include costs of babysitting or travel during the time having medical care. Finally, the cost of time seeking medical care and lost productivity due to illness are other opportunity costs of healthcare interventions that do not result in money exchanging hands, though it truly is a cost of healthcare.

Perform discounting Discounting is the procedure that calibrates outcomes that occur over time. For economic evaluation of healthcare programs, both costs and effectiveness must be appropriately discounted. The Panel on CostEffectiveness in Health and Medicine has recommended that a discount rate of 3% should be used. For example, at a discount rate of 3%, a cost of $1 spent next year is equivalent to the cost of 97 cents spent today. If the costs of an intervention total $1000 each year over a three year period, then the total cost attributed to that program in present dollars should be calculated to be $1000 + $1000/1.03 + $1000/(1.032 ). Over long periods of time, discounting can have a major impact. The cost of $1 spent in ten years’ time is equivalent to spending only 74 cents today.


the model and determining how the bottom line result would change; for example, what would be the changes in the incremental cost-effectiveness ratio if the prevalence of disease would change.

Address ethical issues Ethical issues need to be addressed when performing technology assessment studies. Many situations arise in cost effectiveness analysis in which the best decision for a single patient is not the same decision that would be made when taking into account society as a whole. Specifically, a clinical intervention may be found to be clinically effective, but it might not be cost effective. Thus, what might be best for a single individual might not be best for society as a whole. This happens when an expensive clinical intervention yields minimal benefits. Such a situation is frequent in the oncology literature, where expensive interventions yield marginal health benefits at a substantial cost.

Interpret the results The interpretation of results of an economic evaluation is an important component of the summary of the analysis. Results are often presented as a ratio, where the numerator includes all the costs and the denominator includes all the outcomes. Many times, results are presented as an incremental analysis, comparing the costs and benefits of a new strategy to the next most effective strategy. For example, to compare the effectiveness of the use of an HPV vaccine and screening to that of screening alone one would calculate the following ratio: [costs of vaccination & screening]−[costs of screening] . [outcomes of vaccination & screening]−[outcomes of screening] (11.1)

Perform sensitivity analysis Sensitivity analysis is the process of varying parameters in an analysis to determine if the selected optimal alternative would remain optimal. It is crucial to examine whether the output of model-based cost effectiveness analysis is extremely sensitive to small changes in one or more input parameters. Most frequently, sensitivity analysis is performed by varying a single parameter of

The implications of the results should be stated, with a particular emphasis on how the results may affect clinical practice or public policy. One important issue to address is the issue of magnitude. The economic analysis may yield an incremental cost effectiveness ratio and the important question to answer is whether this ratio is a “big number” or a “little number.” In other words, is


Biomedical Engineering for Global Health

Table 11.1. A league table ranks interventions in terms of their relative cost effectiveness [15]. Intervention

Cost effectiveness ratio

Pneumococcal vaccine for adults over 65 years old

Cost saving

Tobacco cessation counseling

Cost saving to $2,000/QALY saved

Chlamydia screening of women 15–24 years old

$2,500/QALY saved

Colorectal cancer screening for people more than 50 years old

$13,000/QALY saved

the bottom line result significant enough to make an impact or a difference? Should the result imply a change in clinical practice or public policy? Some investigators have chosen to use a “league table” to make comparisons between the results of the cost effectiveness analysis under investigation and the results of previously performed cost effectiveness analyses that have been previously published in the literature (Table 11.1). These league tables (named after the European soccer leagues, which publish the standings or rankings of teams in the newspapers) provide one measure of comparison. The reasoning goes that if the incremental cost effectiveness ratio for a given analysis is similar to the incremental cost effectiveness ratio for a previously performed analysis and the previously analyzed intervention is considered an acceptable clinical practice that is also “cost effective,” then the intervention under investigation should also be considered a cost effective use of resources. Many articles in the clinical evaluation literature have cited $50 000 per quality adjusted life year (QALY) as an appropriate threshold for determining the cost effectiveness of a healthcare intervention. This threshold was determined from earlier studies that have been considered acceptable and a good value (e.g. treatment of moderate hypertension) or acceptable though a controversial value (e.g. dialysis for early stage renal disease). However, there is no specific reason or benchmark that presently applies to indicate that $50,000 per QALY is the threshold for optimal decision making regarding resource allocation.

A lack of resources prevents the application of this rule in many developing countries. In this setting, an alternative that has been suggested is to consider interventions to be very cost effective if the amount that must be spent to gain one QALY is less than the per capita GDP, and cost effective if it is less than three times the per capita GDP [6].

Examples of cost effectiveness analysis With this overview, we next consider several examples of cost effectiveness analysis in detail. In this chapter, we will focus on the problem of screening for cervical cancer. More than 50 million Pap tests are performed annually in the USA, and the costs of screening exceed several billion dollars annually [6]. With such a large investment, it is imperative to ensure that we continue to invest public health resources wisely. With so many new screening technologies available, we will find that cost effectiveness analysis can help us decide which test to use, who should be screened and how frequently they should be screened. Our goal is to maximize clinical benefits for women in the face of competing health problems and limited resources. We will see that the solutions to this trade-off vary substantially as we move from the developed to the developing world. We begin our analysis by going back to 1988 in the USA. At that time, Medicare did not provide coverage for cervical cancer screening, even though the elderly accounted for 40% of cervical cancer deaths. Researchers wanted to determine whether cervical cancer screening would be cost effective in this population [7]. To answer this question, a study was carried out to examine the costs and benefits of screening elderly low income women; researchers adopted a societal perspective for all costs and benefits. The new technology to be assessed was the Pap smear and the alternative considered was no screening. To calculate the effectiveness and the costs of the new technology, researchers combined data from a real clinical trial with projections of future costs and benefits from an economic model. All cost and benefit calculations used a 5% discount rate. The results indicated that the use of Pap smear screening would be cost saving in this group of women;

Cost effectiveness of screening for disease


Projecting costs and benefits into the future: Markov models Cost effectiveness analysis can be carried out using real data from clinical trials or by using mathematical models to project costs and benefits. Since we are usually interested in assessing costs and benefits over a patient’s lifetime, most studies use the model-based approach because it provides a rapid, inexpensive way to compare the impact of many different strategies. Markov modeling is a mathematical technique that allows us to follow a cohort of patients over a defined period of time. In a Markov model we simulate the health of a group of patients over a defined period of time. At short time intervals, we use a probabilistic approach to predict each patient’s health state, and we keep track of the costs and benefits associated with a medical intervention. If we want to compare the costs and benefits of several interventions, we can use a Markov model to simulate the effects of different interventions on several groups of patients, comparing the costs and benefits of each intervention. Markov modeling is a useful tool because it allows us to quickly and economically compare many different strategies; we can use the results of Markov modeling to help guide the design of clinical trials to test interventions that appear to be most cost-effective based on modeling studies. The first step in building a Markov model is to define a cohort of patients, and to define all the different health states that the patients can experience. Let’s take the example of cervical cancer screening. As the starting point for analysis, we can take a cohort of 18-year-old women who initially present for screening of cervical pre-cancer. Using a Markov model, we wish to follow this cohort of women through their lifetimes to determine the expected economic costs and health benefits of using different strategies for the screening, diagnosis and management of cervical pre-cancer. The figure above show the different health states (relevant to cervical cancer) that our patients can experience. During the course of the initial model cycle, women can either develop HPV or not, or die from causes other than cervical cancer. In the simulation, women move through the model, either progressing, persisting, or regressing from health state to health state. The health states include “NORMAL”, “HPV”, “LGSIL”, “HGSIL”, “EICC” (Early-Invasive Cervical Cancer), “LICC” (Late-Invasive Cervical Cancer), and “DEATH”. This Markov process models the natural history of cervical pre-cancer, and in order for our results to be meaningful, we must have reliable information about the probabilities of disease regression and progression for each health state. In order to implement a Markov model, we must choose a time interval to update each patient’s health within the simulation. For analyses of cervical cancer screening, often a cycle length of six months is selected because six months is the length of time that occurs between follow-up visits should a woman either require follow-up after treatment, or require follow-up after a false positive screening test. To compare the cost effectiveness of different interventions using this approach, we can either let women progress through the model with no screening, or we can superimpose screening, diagnosis, and treatment on the model, to reflect the alternatives for evaluation.

sensitivity analysis indicated that this result was true over a broad range of model parameters. The researchers recommended that screening should be made available to previously unscreened elderly women, which raised the ethical dilemma of whether

Medicare should restrict a benefit to only those women who had never before been screened. Let’s examine the details of this analysis. In the study, a group of women aged 65 years and older who were seeking care in a municipal hospital outpatient clinic


Biomedical Engineering for Global Health

were invited to participate in a cervical cancer screening program; 816 women agreed to participate in the study, and 25% of these women had never had a Pap smear. The researchers monitored the number of cervical cancers and pre-cancers in the screened group; they found that 11 women had abnormal Pap smears. Follow up testing confirmed two patients with invasive cancer, two with high grade pre-cancer and three with low grade pre-cancers. They then added up the actual costs of screening and treatment in this group of women which amounted to $59,733. To estimate the benefits of screening, they used a Markov model to predict the number of years of life gained by early detection. They assumed that patients with pre-cancer have a 100% survival rate; using known values of progression rates of cervical pre-cancer and

Summary of cost effectiveness analysis of cervical cancer screening for low income, elderly women [7] New technology:

Alternative: Number of Pap tests performed: Costs of technology: Benefits of technology:


Costs of alternative:

$107,936 to treat invasive cancers that

mortality rates of cervical cancer, they calculated the life expectancy of women if screened and if not screened. The difference in these projections yielded the number of years of life gained through early detection. They found that screening saved a total of 30.33 years of life in

Net costs of intervention:

this group of 816 women. Adjustments for the decrease in quality of life due to cervical cancer were also taken into account, in order to estimate the number of quality

Cost effectiveness:

adjusted life years gained by the intervention. Because the quality of life was better with early detection, the intervention added 36.77 quality adjusted years of life in this group. To estimate the costs associated with the alternative of no screening, researchers used the results of their Markov model to estimate the costs associated with treating the cancers that would have developed in the absence of screening in this same group of women. They estimated that it would have cost $107,936 to treat the same group of women if their cancers had not been identified until they sought treatment because of symptoms. Comparing the cost of the new technology ($59,733) to the alternative of no screening ($107,936), shows that this cervical cancer screening program is COST SAVING! It saves medical costs by early detection of precancer and improves life expectancy. For every 100 Pap tests performed, the program saved 3.72 years of life and generated $5907 in savings by averting the need for future expensive cancer treatments [7].

Pap smear screening in low income elderly women No screening

$59,733 30.33 life years gained 36.77 QALYs gained

develop in the absence of screening $59,733 − $107,936 = −$48,203 (intervention saves money) SAVE $1311/QALY

In part, as a result of this study, Medicare benefits were extended to cover triennial screening with Pap smears in 1990 for all women with no upper age limit [8]. Is this a cost effective use of resources? The study we just considered examined only a one-time screen in a population with limited prior access to screening. Should these results be generalized to an entire population of elderly women? Should there be any upper age limit on screening? These questions were considered in a follow-up study by the same group that showed that programs targeting women who have never been screened will be cost saving and that triennial screening is the most cost effective approach [9]. For a population representative of the USA, where most women have had some screening, triennial screening will cost $2254 per year of life saved, whereas annual screening will cost $7345 per year of life saved. For women who have regularly been screened, costs of annual screening increase to $33,752 per year of life saved. Based on these results,

Cost effectiveness of screening for disease researchers recommend that women over the age of 65 can stop screening if they have a history of negative smears. In Chapter 10, we saw a number of new approaches to improve cervical cancer screening, including liquid cytology and HPV DNA testing. Cost effectiveness analysis can also help us compare the net benefits and costs of adopting these new technologies compared to screening with the conventional Pap test. To compare the cost effectiveness of new interventions for cervical cancer screening, an economic analysis


Table 11.2. Performance and cost of new technologies for cervical cancer screening [11]. Technology



Cost per Test

Liquid Cytology












HPV + cytology




was recently reported. Strategies considered included: (1) the conventional Pap test, (2) liquid based cytology followed by HPV DNA testing in positive cases, (3) HPV DNA testing and conventional Pap in women over the age of 30, and (4) HPV DNA testing and liquid based

absence of screening is only 28.7 years. With screening this rises to 28.78 years. Results are coded according to the frequency of screening: blue and red correspond to

cytology in women over the age of 30. In all cases, the

or every four years). Compared to no screening, screening every four years yields a large gain in life expectancy at a nominal cost. The cost per year of life saved using the conventional Pap smear every four years is $9400

frequency of screening was varied and the alternative was no screening. The costs and performance assumed for each technology are summarized in Table 11.2. A societal perspective was adopted [11]. Results of the analysis are shown in Figure 11.2. For each strategy, the figure shows the average life expectancy as a function of the per person total costs. Both costs and life years were discounted at an annual rate of 3%. Because life years have been discounted at this rate, the total discounted life expectancy in the

more frequent screening (annual or biennial) and black and yellow represent less frequent screening (triennial

(Point 1). We see that as the frequency of screening increases, there are small gains in life expectancy, but large gains in cost. What is the marginal cost of these additional gains? To assess the incremental cost effectiveness, we divide the increase in life expectancy by the increase in cost. For example, as we go from screening with the

Figure 11.2. Discounted life expectancy versus per person costs of cervical cancer screening for several different technologies and screening intervals [11].


Biomedical Engineering for Global Health

Impact of cost effectiveness study of cervical cancer screening for low income, elderly women Before 1989, Medicare did not provide coverage for routine Pap tests. Representative Cardis Collins of Illinois introduced a bill to add coverage for Pap smears to Medicare every year for 15 years before this benefit was added in 1989. The economic analysis carried out by Jeanne Mandelblatt and colleagues was very influential in convincing legislators to vote for the bill. Dr. Mandelblatt recalls: “I previously worked in the Harlem community and other New York City neighborhoods that were very poor in resources: housing, healthcare, and other resources. The issue I wanted to address was whether we should screen older women for cervical cancer. The reason I, as opposed to someone else, did this is that I was the only person in the primary care clinic who knew how to do gynecologic examinations, and I was the first person in 10 years to observe that the examination tables had stirrups! This was the beginning of my life’s work. In the first few years of our screening program, the nurse practitioner and I screened more than 800 women. They were on average 74 years old and had Courtesy of Jeanne Mandleblatt, MD, MPH, Georgetown University Medical Center.

largely been unscreened previously. As a result, we found that screening these women actually saved lives as well as healthcare costs (3.72 lives and $5,907

saved for every 100 Pap smears done)-an ideal program. But then serendipity came into play. We were doing this work at a time when there was an explosion in the growth of the older population and members of Congress were receiving a lot of pressure from their older constituents to include preventive services. Along I came with my Pap smear analysis and showed that if we were to screen the average elderly population at that point Pap smear screening would be a good buy. It would cost about $2,200 per year of life saved. Of great importance was that we could save money if we targeted screening to women who had not been screened previously, but the cost-effectiveness would worsen by more than 10-fold if screening were applied to women who had already been regularly screened. What were our responsibilities and what were the issues that came out of this work? When we presented this work to the Office of Technology Assessment, we proposed considering cervical cancer screening as a targeted benefit and perhaps even including benefits to do outreach to women who have never been screened. The OTA said that under Medicare, benefits must be included for all (or no) women, so our recommendation could not be implemented. . . . The actual cost-effectiveness for Medicare might not be as favorable as it could have been if targeted to the highest-risk women” [10].

conventional Pap every four years (Point 1) to screening with liquid cytology followed by HPV DNA testing (Point 2), we must spend $20,600 for every additional year of life gained. This incremental cost effectiveness is simply the inverse of the slope of the line connecting two

cytology and HPV DNA testing every two years (Point 3) with annual testing (Point 4), we see that we must spend an additional $2,215,100 for every year of life gained – not a cost effective use of resources. Let’s use this graph to compare the cost effectiveness

adjacent points on the graph. As the slope approaches zero (horizontal line), the incremental cost effectiveness decreases. For example, if we compare screening women over the age of 30 with a combination of liquid

of an annual conventional Pap smear (Point 5) with less frequent use of liquid cytology and HPV DNA testing every 2 years (Point 3). We see that the model predicts we will save more lives and spend less money with

Cost effectiveness of screening for disease this new strategy. This type of analysis can be used by professional societies to help make policy recommendations about screening strategies. We can use a similar strategy to make recommendations about screening in the developing world. Researchers compared the cost effectiveness of different strategies for cervical cancer screening in five developing countries: India, Kenya, Peru, South Africa, and Thailand [12]. The strategies considered included direct visual inspection with acetic acid (DVI – also called VIA), conventional Pap testing, and HPV DNA

Table 11.3. Performance of cervical cancer screening strategies considered in cost effectiveness analysis in developing countries [12]. Intervention












Discounted life expectancy According to the CDC, the life expectancy at birth for women in the United States in 2004 is 80 years. If we discount the value of future life by 3% per

testing, compared to the alternative of no screening (Table 11.3). The number of clinic visits required was also varied including strategies that required separate visits for screening, colposcopic diagnosis, and treatment (three visit strategies), and separate visits for

year, what is the equivalent discounted life expectancy at birth? To make this calculation, we simply add up the present value of all future years of life, discounting the value of future life years by 3% per year.

screening and treatment (two visit strategies). In the case of DVI, results are available immediately, so that a one visit strategy with screening and treatment on the same day was also considered. In each case, the frequency of screening was varied, including screening

Discounted life expectancy = 1 + 1/1.03 + 1/(1.03)2 + · · · + 1/(1.03)79

once, twice, or three times within a woman’s life, and screening every five years. Table 11.4 compares the demographic and economic characteristics of the five countries. The costs of clinical care were based on local labor costs. Despite wide variations in the incidence of cervical cancer and the availability of resources, results indicated that one or two

You can easily use a spreadsheet to show that the discounted life expectancy is only 31.1 years.

visit screenings with DVI or HPV DNA testing once per lifetime reduced the lifetime risk of cancer by approximately 25–36% at a cost of less than $500 per year of

Table 11.4. Demographic and economic characteristics of five developing countries considered in a c 2005 Massachussetts Medical Society. study of cost effectiveness of cervical cancer screening [12].  All rights reserved.

Demographic and Economic Characteristics of Five Developing Countries India


Total population (millions)






Rural population (% of total)






Population density (# of persons/km2) Women 35–39 yr of age (% of total pop.)

341.69 3.28

52.87 2.18

20.26 3.21

36.03 3.35

118.87 4.1

Literacy rate among women ≥15 yr of age (%)











Average hourly wage rate (2000 international dollars)






Female life expectancy at birth (yrs)






Cervical cancer incidence (age-standardized inc./100,000)§
















Women employed in informal sector (% employed women)

HIV prevalence among adults (% of total pop.) Per capita GDP (2000 international dollars)



South Africa Thailand


Biomedical Engineering for Global Health

Figure 11.3. Discounted life expectancy versus discounted total lifetime costs of screening for cervical cancer in South c 2001, Africa from [13]. Copyright  American Medical Association.

life saved, a figure that is less than the per person GDP in each of the five countries [12]. Let’s examine the results from South Africa in more detail. Figure 11.3 shows the results of the analysis, graphing discounted life expectancy versus discounted total lifetime costs per patient. Note that the base discounted life expectancy in South Africa is only 19 years. Single visit strategies based on DVI are shown as black circles, two visit strategies based on HPV testing are shown as gray circles and three visit strategies based on cytology are shown as open circles. In general, single visit strategies save more lives and cost less than multi-visit strategies. As the frequency of screening is increased, more lives are saved but at higher costs. The incremental increase in cost effectiveness jumps from $140 per year of life saved for screening with DVI every five years to $460 per year of life saved when DVI is used every three years [13].

Summary What can cost effectiveness analysis tell us about the best use of resources for cervical cancer screening? In developed countries, cost effectiveness decreases as the frequency of screening increases to more than every two or three years. Strategies that improve sensitivity (liqud based cytology, HPV testing) offer little benefit at drastically increased costs, unless the screening interval

is increased. When the screening interval is increased to three to four years with these tests, results can be extremely cost effective. Small changes in specificity have a large influence on the cost effectiveness of strategies that involve frequent screening. Screening after age 65 is not cost effective for women who have had consistently negative screening results. For older women who have had no prior screening, initiating screening is very cost effective [6]. In the developing world, screening efforts should target women age 35 or older and should screen all women at least once in their lives. If a large fraction of women can be screened, then screening two to three times per lifetime could reduce lifetime cancer risk by 25–40%. If three lifetime screens are offered, efforts should target women aged 30–50, with a screen about every five years [6]. It is interesting to directly compare the cost effectiveness of screening in the developing and developed world. Figure 11.4 shows the reduction in lifetime risk of developing cervical cancer as a function of the lifetime costs per person for cervical cancer screening in South Africa and the United States. In the lower left region of the graph, increasing the frequency of screening in South Africa from once to twice to three times in a woman’s lifetime shows a region of rapidly escalating benefits at low incremental cost relative to doing nothing. Three strategies from the USA (triennial,

Cost effectiveness of screening for disease


Figure 11.4. Reduction in lifetime risk of developing cervical cancer versus total lifetime costs of screening. The graph compares strategies in a developing country (South Africa) to that of a developed country (United States). Used with permission from [14]. This article was published in Virus Research, Vol. 89, S. J. Goldie, Health economics and cervical cancer prevention, pp. 301–9, copyright Elsevier (2002).

illustrates the very different challenges facing devel-

Bioengineering and Global Health Project Project task 7: Develop an alternative solution that meets the constraints you specified This could be either an improvement to an existing device or a completely new strategy. For example, if your project is to develop alternative TB therapies, you might propose an alternate drug delivery mechanism (e.g. patches, controlled-release injections, implantable devices, etc.) that could be used to deliver six to nine months of TB therapy without requiring frequent physician visits or daily drugs. Remember, improvements are not limited only to changes in device design, but may also include increasing availability of the device to healthcare workers in the field or decreasing the cost to manufacture and distribute a health technology. Turn in a one-page summary of your new solution. The summary should include the following: (1) overview of your design, (2) scientific principles of the design, (3) expected benefits of the design, (4) potential risks associated with the design, (5) a plan to determine whether the design meets the constraints specified in Task 6.

biennial and annual screening) are on a region of the curve where the slope is quite flat and there are very small incremental benefits at high incremental costs associated with more frequent screening. This graph

oped and developing countries. The challenge for the USA is to ensure that the price we pay for achieving small gains in life expectancy does not increase disparities in access to care. In contrast, the challenge for developing countries is to simply get on the curve. The extreme disparity in cost effectiveness can be illustrated by comparing how much additional life one can buy for an investment of $50,000. In the USA, an additional investment of $50,000 toward cervical cancer screening can buy 15 weeks of life. In South Africa, this same investment can buy more than 1000 years of additional life [14]! Thus, the process of health technology assessment is critical for decision making in all countries. It can be used to help answer both clinical and policy questions and should be performed at all stages of the technology development process, to ensure appropriate and efficient allocation of scarce healthcare dollars.

Homework 1. What is a QALY? How much is our society willing to spend to gain one QALY? 2. Draw and label a graphical representation of health policy space. In which quadrant of the graph would the following interventions be located? a. Measles vaccinations for children. b. Anti-retroviral drug therapies for HIV infected patients.


Biomedical Engineering for Global Health

c. Screening all women for ovarian cancer using serum CA-125 levels. 3. Explain why it is important to consider “additive procedures” when analyzing the cost effectiveness of new medical technologies. Give a specific example in which additive procedures influenced cost effectiveness. 4. Cardiovascular disease is the leading cause of death in the United States. a. What are four major treatments for coronary artery disease? b. Currently, coronary artery disease is diagnosed using coronary angiography, a painful and expensive technique. You have just developed a new, painless technique to diagnose coronary artery disease. The technique is also substantially

c. If the new test was used, how many years of life would be gained? d. If this test was administered annually to all women over age 40, how many $ would we spend per year of life gained? e. Based on your answer to part d, do you think this test would be adopted in the developed world? In the developing world? Explain your reasoning. 6. In the USA, HIV infection is often discovered at an advanced state, when patients seek treatment for complications due to AIDS. When detected early, HIV infection can be controlled using HAART. On average, early detection of HIV infection extends the life of HIV patients by 1.8 years. Recent articles in the February 10, 2005, issue of the New England

less expensive than coronary angiography.

Journal of Medicine argue that we should consider

Somewhat surprisingly, some health economists predict that the introduction of your new technique will actually cause healthcare expenditures to grow. Why might they make this

voluntary screening of all patients, particularly those in higher risk groups. In 2006, the CDC recommended that people aged 13–64 should be screened for HIV regardless of risk factors. In this

prediction? 5. When detected early, ovarian cancer is curable for 95% of women. Unfortunately, in the majority of

problem, you will be asked to do some calculations to determine whether you agree with their conclusions. The prevalence of HIV infection in the

cases, ovarian cancer is not detected until widespread metastasis has occurred. In this circumstance, ovarian cancer is fatal approximately 63% of the time. The American Cancer Society

general US population is 0.33%. There are 292 million Americans, 140 million of whom are over age 40. The cost of screening for HIV is approximately $2.

estimates that there will be about 25,580 new cases of ovarian cancer in the United States in 2004. About 16,090 American women will die of the disease in 2004. On average, 22 years of life are lost when a woman dies of ovarian cancer. You have developed a new blood test that can detect ovarian cancer in the earliest possible stages. Each blood test costs $200 to perform. There are 292 million

a. How many Americans are infected with HIV? b. If the test was administered to all Americans, how much would we spend in testing? c. If the new test was used, how many years of life would be gained? d. If this test was administered to all Americans, how many $ would be spent per year of life gained?

Americans, approximately 70 million of whom are women over age 40. a. How much money would we spend annually if all women over age 40 were screened with this new test? b. Calculate the annual mortality rate of ovarian

e. Based on your answer to part d, do you think this test would be adopted in the developed world? In the developing world? Explain your reasoning. f. Approximately one person per 200,000 individuals tested will have a false positive test result. In other words, their HIV test will be

cancer without the use of the new test. Compare this to the expected annual mortality rate of ovarian cancer with the use of the new test.

positive even though they do not have HIV. If we test all Americans, how many people will receive a false positive result?

Cost effectiveness of screening for disease 7. You have developed a new technology that could detect pre-cancerous cells in the sputum. This technology can enable much earlier detection of lung cancer, reducing the fraction of lung cancer patients that die of their disease from 90% to 15%. Your test costs $100 to perform. Assume that on average, 18 years of life are lost when a person dies of lung cancer. There are 292 million Americans, 140 million of whom are over age 40. There are 173,770 new cases of lung cancer in the United States each year. a. How much money would we spend annually if all adults over age 40 were screened with this new test? b. Calculate the mortality rate of lung cancer without the use of the new test. Compare this to the expected mortality rate of lung cancer with the use of the new test. c. If the new test was used, how many years of life would be gained? d. If this test was administered annually to all adults over age 40, how many $ would we spend per year of life gained? e. Based on your answer to part d, do you think this test would be adopted in the developed world? In the developing world? Explain your reasoning.

References [1] Littenberg B. Technology assessment in medicine. Academic Medicine. 1992 Jul; 67(7): 424–8. [2] Drummond MF. Methods for the Economic Evaluation of Health Care Programmes. 2nd edn. Oxford: Oxford University Press; 1997. [3] Cantor SB, Ganiats TG. Incremental cost-effectiveness analysis: the optimal strategy depends on the strategy set. Journal of Clinical Epidemiology. 1999 Jun; 52(6): 517–22. [4] Bergus GR, Cantor SB, Ebell MH, Ganiats TG, Glasziou PP, Hagen MD, et al. A glossary of medical decisionmaking terms. Primary Care. 1995 Jun; 22(2): 385–93.


[5] Cantor SB. Pharmacoeconomics of coxib therapy. Journal of Pain and Symptom Management. 2002 Jul; 24(1 Suppl.): S28–37. [6] Goldie S. A public health approach to cervical cancer control: considerations of screening and vaccination strategies. International Journal of Gynaecology and Obstetrics: the official organ of the International Federation of Gynaecology and Obstetrics. 2006 Nov; 94 Suppl. 1: S95–105. [7] Mandelblatt JS, Fahs MC. The cost-effectiveness of cervical cancer screening for low-income elderly women. Jama. 1988 Apr 22–29; 259(16): 2409–13. [8] Power EJ. From the Congressional Office of Technology Assessment. Jama. 1990 Jun 13; 263(22): 2996. [9] Fahs MC, Mandelblatt J, Schechter C, Muller C. Cost effectiveness of cervical cancer screening for the elderly. Annals of Internal Medicine. 1992 Sep 15; 117(6): 520–7. [10] Hagen MD, Garber AM, Goldie SJ, Lafata JE, Mandelblatt J, Meltzer D, et al. Does cost-effectiveness analysis make a difference? Lessons from Pap smears. Symposium. Medical Decision Making. 2001 Jul–Aug; 21(4): 307–23. [11] Goldie SJ, Kim JJ, Wright TC. Cost-effectiveness of human papillomavirus DNA testing for cervical cancer screening in women aged 30 years or more. Obstetrics and Gynecology. 2004 Apr; 103(4): 619–31. [12] Goldie SJ, Gaffikin L, Goldhaber-Fiebert JD, Gordillo-Tobar A, Levin C, Mahe C, et al. Cost-effectiveness of cervical-cancer screening in five developing countries. The New England Journal of Medicine. 2005 Nov 17; 353(20): 2158–68. [13] Goldie SJ, Kuhn L, Denny L, Pollack A, Wright TC. Policy analysis of cervical cancer screening strategies in low-resource settings: clinical benefits and cost-effectiveness. Jama. 2001 Jun 27; 285(24): 3107–15. [14] Goldie SJ. Health economics and cervical cancer prevention: a global perspective. Virus Research. 2002 Nov; 89(2): 301–9. [15] Salinsky, E. “Clinical Preventive Services: When is the Juice Worth the Squeeze?” National Health Policy Forum Issue Brief. 2005; 806.

12 Technologies for treatment of heart disease

We have examined how new technologies can be used to prevent infectious disease. When prevention fails, we need effective methods to both detect and treat disease. In Chapter 10, we showed that new imaging technologies can play a critical role in improving the early detection of cancer. By identifying cancers and pre-cancers at an early stage, we provide an opportunity to intervene when the disease is most responsive to therapy. We now examine the use of technology to treat disease. Our focus is on cardiovascular disease, where the development of medical therapies, new surgical procedures and implantable cardiac devices (Figure 12.1) has led to dramatic reductions in cardiovascular mortality in developed countries over the past 50 years. Despite these advances, we will see that cardiovascular disease is still the leading cause of death in the United States. Moreover, the mortality of cardiovascular disease is rapidly increasing in developing countries. We will see that future advances in cardiovascular devices such as drug eluting stents, surgical robots, and implantable artificial hearts have the promise to further reduce cardiovascular mortality. However, these technologies are expensive and require infrastructure beyond that which is currently available in most developing countries. More cost effective treatments, together with a greater emphasis on the prevention of cardiovascular disease, are needed to address this growing burden of disease.

The global burden of cardiovascular disease Almost 25% of the USA population – 61 million Americans – suffers from cardiovascular disease. Today, heart disease accounts for more than 40% of all deaths in the USA, with 950,000 Americans dying of heart disease every year. The costs of cardiovascular disease in the USA exceed $350 billion per year. More than $200 billion is spent annually on direct healthcare expenditures and $140 billion in productivity is lost each year due to premature death and disability. Globally, cardiovascular disease was responsible for 16.6 million deaths in 2001, representing nearly 1/3 of global mortality. More than 80% of deaths due to cardiovascular disease occur in low and middle income countries, where access to interventional therapies is least available. By 2010, cardiovascular disease is predicted to be the leading cause of death in developing countries. Heart disease is also an important cause of global morbidity. Every year, more than 20 million people survive heart attacks and strokes, and the cost of caring for these patients is high [1]. The changing patterns of incidence of cardiovascular disease present a unique global challenge (Figure 12.2). At the beginning of the twentieth century, cardiovascular disease accounted for less than 10% of all deaths worldwide; today it accounts for nearly half of mortality in the developed world and one quarter of mortality

Technologies for treatment of heart disease




Figure 12.2. Unhealthy lifestyles put billions of people at risk for cardiovascular disease and sudden death. World Health Organization, Integrated Management of Cardiovascular Risk: Report of a WHO Meeting, 2002.

to suffer from the chronic diseases of middle and old age. However, the increasing global mortality of cardiovascular disease also partly reflects changes in lifestyle. The risk factors for developing cardiovascular disease include tobacco use, low levels of physical activity,

Figure 12.1. (a) The use of small implantable cardiac devices which can be temporarily or permanently placed in the chest cavity can assist patients suffering from end stage heart failure. (b) The Jarvik 2000, is an example of this new type of device which assists the left ventricle in pumping oxygenated blood to the rest of the body. R Institute. Texas Heart

in the developing world [2]. The increasing incidence and mortality associated with cardiovascular disease is partly due to improvements in public health which have reduced the incidence and mortality of infectious disease; as people live longer lives, they are more likely

inappropriate diet, high blood pressure, and high serum cholesterol levels. The rate of increase in the number of persons at risk for cardiovascular disease is most pronounced in developing countries. In Chapter 4, we considered the epidemiologic transition that is occurring in many developing countries. We saw that, as societies develop and life expectancy increases, the causes of death shift from primarily infectious disease to chronic diseases [2]. This shift is driven by a combination of factors, including changes in diet, patterns of physical activity, and tobacco consumption [1]. Food market globalization and the increased availability of cheap, higher fat foods lead to increased caloric intake, while mechanization and urbanization lead to decreased caloric expenditure [2]. Resultant increases in the underlying risk factors of


Biomedical Engineering for Global Health income people consuming a high fat diet rose from 23%

500 Total Cardiovascular Disease


400 300 Diseases of the Heart

200 Coronary Heart Disease



0 1900







Figure 12.3. Changing mortality rates of heart disease per 100,000 population throughout the epidemiologic transition in the United States [4].

cardiovascular disease, such as obesity, serum cholesterol level, hypertension and diabetes, drive increases in the incidence of cardiovascular disease. In the USA, this epidemiologic transition took place in the first half of the twentieth century and the incidence of heart disease increased during this period (Figure 12.3). In the latter

to 67% from 1989 to 1993 [3]. As a result, the increase in cardiovascular disease incidence is outpacing the availability of effective treatments that normally accompany the economic growth of the epidemiologic shift. Left unchecked, cardiovascular disease has the potential to severely reduce economic growth. Because cardiovascular disease strikes people in their mid-life years the economic impact is especially severe [1]. If the head of a household dies at an early age due to cardiovascular disease, there is a devastating impact on the entire family [2]. For example, in Bangladesh, a dependent child who loses an adult parent or guardian is more than 12 times likely to die themselves [2]. Of great concern is that fact that people in developing countries tend to die at an earlier age as a result of cardiovascular disease than do people in developed countries [3]. In developed countries, only one quarter of deaths due to cardiovascular disease occur in people under the age of 70. In contrast, nearly half of deaths due to cardiovascular disease in developing countries occur in people under the age of 70 years.

half of the twentieth century, advances in treatment of heart disease led to declines in the mortality of cardiovascular disease. From 1965 to 1990, mortality due to heart disease fell by 50% in Australia, Canada, France, and the USA, and fell by 60% in Japan [3]. Today, the epidemiologic shift that occurred in the USA in the early 1900s is being played out in many developing countries. Because of the global availability of processed foods, developing countries are entering a period of increased incidence of cardiovascular disease sooner in the epidemiologic transition than occurred in the USA. For example, in China the number of upper

Types of CVD Recall from Chapter 4 that there are two main forms of cardiovascular disease (CVD): ischemic heart disease and cerebrovascular disease (stroke). Ischemic heart disease can cause heart attack, and is the leading cause of death in the USA. It is also the leading cause of premature, permanent disability among working adults. Globally, ischemic heart disease is the second leading cause of death in adults aged 15–59 and the leading cause of death among adults over the age of 60 (Table 12.1).

Table 12.1. Top ten causes of global mortality among adults aged 15–59 years and 60 + [5].

Technologies for treatment of heart disease Stroke is the third leading cause of death in the USA, and the second leading cause of death worldwide in adults over the age of 60. Frequently, cardiovascular disease results in sudden death: 1.3 million Americans experienced a heart attack in 2006, and 40% of those who have a heart attack will die from their disease, and half of these deaths occur within one hour of symptom onset, before people reach the hospital [6]. Because so many of these deaths occur without warning, it is important to focus on opportunities to either prevent cardiovascular disease or to diagnose and treat ischemic heart disease before a heart attack occurs.

Opportunities to prevent CVD


income countries, but rates of detection are much lower in developing countries. However, although we detect most cases of high blood pressure in developed countries, it is frequently not treated adequately. Of those with high blood pressure in the USA, over 70% have uncontrolled hypertension (Figure 12.4). For those on treatment, only 13–29% have their hypertension adequately controlled. In African countries, control rates are only 2% [7]. Another screening tool is to monitor serum cholesterol levels every five years. Table 12.2 shows total cholesterol levels considered desirable, borderline and abnormal. A 10% drop in total cholesterol can reduce the risk of heart attack by 30%. There are two types

Modifiable risk factors for developing cardiovascular disease (smoking, unhealthy diet, physical inactivity)

of lipoproteins which carry cholesterol in the blood: LDL and HDL. Cholesterol carried by LDL, or low density lipoprotein, contributes to the development of

lead to symptoms of hypertension, diabetes, obesity, and high serum cholesterol [7]. More than 50% of death and disability due to cardiovascular disease could the-

atherosclerosis. Cholesterol carried by HDL, or high density lipoprotein, seems to protect against heart attack, possibly by transporting to the liver to be safely

oretically be prevented through efforts to reduce high blood pressure, high cholesterol, obesity and smoking [1]. In particular the importance of smoking cessation programs could be enormous. In 1995, there were

excreted. Unfortunately, 80% of patients with high cholesterol levels do not have them adequately under control.

1.1 billion smokers in the world, and tobacco contributed to three million deaths. In 2001, tobacco contributed to five million deaths, and this is expected to rise to ten million by 2020. We can screen for those at risk for developing cardiovascular disease by measuring blood pressure annually: 15–37% of the world’s population suffers from hypertension [7]. Hypertension is considered to be a blood pressure above 140/90 mm Hg, while pre-hypertension to be a blood pressure that stays between 120–139/80– 89 mm Hg, and normal blood pressure should be 120 mm Hg

Inflatable cuff

Pressure gauge

(c) Cuff pressure between 80 and 120 mm Hg


(d) Cuff pressure 30 days, serious non-cardiac disease, pregnancy, psychiatric illness (including drug or alcohol abuse), and an inadequate social support system. The endpoints of the clinical trial were: all-cause mortality through 60 days, and quality of life assessments at 30-day intervals until death. The number of patients initially approved was five, with the possibility to expand to 15 patients in increments of five if the 60-day experience was satisfactory to the FDA.

cal and hospital costs for heart transplant are at least $500,000 [18]. It is predicted that the AbioCor price will drop to about $25,000. If this device can give people five years of life, most experts agree that it will likely be adopted.

Left ventricular assist devices Most patients with heart failure have failure of the left ventricle. Rather than replacing the entire heart, an alternative approach is to implant a pump to assist the weakened left ventricle. Such a device is called a left ventricular assist device (LVAD), and it consists of a small pump which is placed inside the left ventricle. The output of the pump is connected to the ascending aorta (Figure 12.19a,b). The LVAD acts to partially unload the left ventricle, and can slow the progression of heart failure and may even allow for some myocardial recovery. The LVAD can be implanted without the need to remove the patient’s heart. In 1994, the FDA approved the use of the LVAD as a bridge to transplantation for those patients with end stage heart failure who were awaiting a donor heart. Self-contained LVADs were approved in 1998 [18]. Short term use of an LVAD can improve the

Technologies for treatment of heart disease




(b) Implanted TET

Thoracic Unit

Figure 12.18. (cont.)

health of patients awaiting transplantation. In addition, it enables patients with end stage heart failure to remain in their homes with skilled nursing care (at a cost of $50/day) rather than needing to remain in the ICU at a cost of $5000/day [24].

Implanted Battery

Implanted Controller

The AbioCor System has four main parts that are implanted inside the body.

Figure 12.18. (a) Photograph of the AbioCor totally implantable artificial heart. Reprinted with permission from Abiomed, Inc. (b) Model of the AbioCor device implanted in the chest cavity. Reprinted with permission from Abiomed, Inc. (c) Robert Tools, the first recipient of the AbioCor totally implantable artificial heart. Courtesy of John Lair, Jewish Hospital, University of Louisville Health Sciences.

There are two main types of LVADs: positive displacement pulsatile pumps and rotary continuous flow pumps (Table 12.6) [25]. The LVADS initially approved for clinical use were pulsatile pumps that mimic the cycle of diastole and systole in the heart. These LVADS contain mechanical parts that can wear out, such as ball bearings, diaphragms or valves [28]. The next generation of LVADs were based on pumps that deliver continuous flow (Figure 12.20). The advantage of this approach is that continuous flow pumps have fewer mechanical parts, are smaller in size, are quieter to operate, require less power to operate, and are less expensive compared to pulsatile pumps [28, 29]. Despite these advantages, there is debate about whether pulsatile flow is critical to maintain organ function. Patients who receive continuous flow LVADs do not


Biomedical Engineering for Global Health


Table 12.6. The two main categories of LVADs and some examples of each [29]. Pulsatile flow pumps

Continuous flow pumps (rotary pumps)

Positive displacement pumps

Axial pumps

Centrifugal (radial) pumps

• HeartMate

• DeBakey

• VentrAssist

• Novacor

• Jarvik 2000

• DuraHeart

• LionHeart

• HeartMate II

• HeartQuest

• SynCardia TAH

• Incor

• HeartMateIII

• AbioCor TAH

• Gyro Pump

From [29]. Reprinted with permission from the Journal of Artificial Organs.

have a pulse and you can’t measure their blood pressure. On the other hand, pulsatile pumps fail more fre-


quently and they are larger than axial flow pumps [25, 27]. Given the success of LVADs as a temporary bridge to transplantation, researchers have begun to study whether LVADs can be implanted permanently as a treatment for heart failure. This use of LVADs is referred to as destination therapy. A recent clinical trial compared the use of LVADs to medical therapy in patients who were not eligible for cardiac transplant. The REMATCH study (Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure) randomized patients to receive an LVAD or medical therapy. This study used the HeartMate LVAD which required an external connection for power (shown in Figure 12.19). Patients receiving an LVAD had both improved survival (Figure 12.21) and

Figure 12.19. Schematic diagram (a) and photo (b) of the HeartMate LVAD. (a) Rose et al. (2001). Long-term use of a ventricular assist device for end stage heart failure. NEJM. 345(20): c 2001 Massachusetts Medical Society. All 1435–43. Copyright  rights reserved. (b) Reprinted with permission from Thoratec Corporation.

quality of life. At one year, 48% of LVAD patients were still alive, compared to only 26% of those receiving medical therapy. At two years, the survival rate in the LVAD group was 26%, compared to 8% for medical management [19]. However, patients receiving the LVAD had higher complication rates, with 28% of patients developing infection at the device site by three months and 42% of patients experiencing bleeding by six months, and a 35% device failure rate at two years [18]: 41% of

Technologies for treatment of heart disease



(b) 100

Survival (%)

80 60

LV assist device

40 20 Medical therapy 0 0






5 3

1 0

Months No. AT RISK

Figure 12.20. There are currently two types of continuous flow pumps in development: centrifugal and axial flow pumps. The Micromed DeBakaey VAD shown here is an example of an axial flow pump [29]. Courtesy of This device was developed jointly between NASA, Baylor and MicroMed Technology Inc. The LVAD is only 25 mm in diameter and 75 mm in length and requires less than 10 W of input power. The axial flow pump generates a rotational speed of 10,000 rpm and can deliver flow rates of up to 10 l/min. From [25].

deaths in the LVAD group were due to sepsis, and 17% of deaths in the LVAD group were due to device failure indicating the challenges associated with mechanical circulatory assist devices [29]. Currently, there are three ways of thinking about the future clinical uses for LVADS. (1) They can be used as a bridge to recovery in patients who have heart failure. Temporary implantation of an LVAD can provide time for the myocardium to remodel and heal [30]. (2) They can be used as a bridge to transplant for those patients in end stage heart failure, sustaining them until a donor heart can be found or an artificial heart is available. (3) They can be used as destination therapy for patients

LV assist device Medical therapy

68 61

38 27

22 11

11 4

Figure 12.21. Results of REMATCH clinical trial. Rose et al. (2001). Long-term use of a ventricular assist device for end stage heart failure. c 2001 Massachusetts Medical Society. NEJM. 345(20): 1435–43.  All rights reserved.

with end stage heart failure. As of 2005, the FDA had approved devices for the last two indications [19]. Based on the available clinical data, it appears that patients do better if LVADs are implanted early, before multi-organ failure begins. As a result of these trials, LVADs have evolved more towards being permanent devices. One challenge for the future is that as more people awaiting transplantation are sustained for longer periods of time on LVADs, the backlog of patients awaiting heart transplants will grow. In order to be successfully used as destination therapy, future LVADS will need to be substantially more durable [31].

Prevention of cardiovascular disease We have seen the economic, technical and clinical challenges of treating end stage heart disease. Yet, the risk factors associated with heart disease are well known


Biomedical Engineering for Global Health

HeartMate III (a)

One continuous flow LVAD under development is the HeartMate III. The pump has no mechanical bearings, but instead uses a magnetically suspended rotor which eliminates contact between moving parts and extends wear life [28]. The device has an upper housing with inflow and outflow channels and a lower housing that contains the motor and rotor. The device diameter is 69 mm, and the height is 30 mm. It uses transcutaneous energy transmission to provide power. The device is made of titanium except for polyester grafts which connect to the patient’s heart and aorta. The titanium surfaces are textured to promote the development of a layer of biological material at the surface upon contact with blood. This pseudo-intima reduces the need for anticoagulation and reduces risk of thrombus development. Blood flows through the device in a circular pattern [28]. The figures in this box are reproduced from Bourque, Kevin, Gernes, David B., HeartMate III: pump design for a centrifugal LVAD with a magnetically levitated rotor. ASAIO Journal. Volume 47(4), July 2001, pp. 401–5. (b)

and many can be avoided or reduced through changes in lifestyle. There is growing awareness of the benefits of an increased emphasis on prevention of cardiovascular disease, particularly in developing countries facing a

There are successful examples of governmental policies to combat heart disease. For example, the Polish government cut subsidies for and imposed taxes to raise the price of animal fats. In response, people cut their

future epidemic of heart disease.

consumption of animal fats and used more vegetable

Technologies for treatment of heart disease oil. The result was a decrease in cardiovascular disease mortality in the 1990s [32]. Is prevention cost effective? We spend more than $100 B in the USA on coronary heart disease. Studies suggest that educational programs dedicated to smoking reduction cost less than $500 per year of life saved, and may even be cost saving [33]. Exercise programs probably cost $10,000 per year of life saved and less than $5000 per QALY saved. For people who actually enjoy exercise, then the health benefits of exercise measured in QALYs may be cost saving! Screening for hypertension and treating with drugs costs about $20,000 per year of life saved. Despite the potential impact of prevention policies, it is frequently more difficult to get people galvanized around prevention efforts in developing countries. In developed countries, public health efforts commenced when the epidemic of cardiovascular disease was at its peak and there were few treatments available [3]. As a result, the public was receptive to these activities. In contrast, developing countries are grappling with the double burden of infectious and chronic diseases and community awareness of the dangers of cardiovascular disease is not high. When implemented, prevention too often focuses on a single risk factor. However, it is much more effective to focus on integrated management of risk: for example, by simultaneously addressing hypertension, smoking cessation and diabetes. A barrier to implementation of such approaches in low and middle income countries is the lack of infrastructure. For example, the lack of basic equipment such as blood pressure measuring devices can make implementing programs to control hypertension impossible. In Nigeria, only 10% of facilities could measure blood lipid levels and 11% of healthcare facilities did not have devices to measure blood pressure. A recent study of hypertension patients showed that 98% of patients have to pay out of pocket for prescriptions. This cost burden is mainly borne by patients. The inability to pay for drugs was cited as a major reason for not taking prescriptions. In Pakistan, the cost of a monthly prescription for hypertension medications ranges from 50% to 200% of monthly per capita income. In moving forward to combat heart disease, in


many ways, the challenge is simply to “use what we know [7].”

Bioengineering and Global Health Project Project task 8: Construct a prototype for an in-class demonstration and design review The audience for your presentation will be a scientific review panel from the World Health Organization. The panel will review the proposed designs and decide which efforts will receive funding to move into the development phase. Therefore, you must convince the panel of the severity of the problem as well as the efficacy of the proposed design. Your prototype does not need to actually function, but it should illustrate the scientific principles of the device as well as the ways in which it satisfies the design requirements. In constructing your prototype, we suggest that you use common household materials (plastic wrap, aluminum foil, cardboard, etc.) to illustrate how the device would work. You are limited to a total materials cost of $10. You should be prepared for questions from the review panel following your presentation.

Homework 1. When a sphygmomanometer is used to measure blood pressure, what two values are measured? Describe how the sphygmomanometer is used to perform these measurements. 2. If a person’s cardiac output is 5.5 l/minute at a heart rate of 75 beats/minute, what is the stroke volume in ml? How many seconds will it take for this person’s entire 4.9 l volume of blood to be pumped through the heart? 3. An avid runner is out for her morning jog before classes. a. Before she starts her jog she measures her pulse for 15 seconds. She counts about 17 beats in the 15 seconds. What is her heart rate in beats per minute?


Biomedical Engineering for Global Health b. This same jogger buys a heart rate monitor because she wants to see her heart rate through different parts of her run. When she is in the

diseases in 2004: white males, black males, white females and black females. c. How do smoking cessation, weight loss, and

middle of her workout the monitor reads her heart rate as 119 bpm. If she were to feel her pulse at this same time for 15 seconds about how many beats would she feel? c. Assuming that her heart stroke volume is about 70 ml, what would her cardiac output be in both parts a and b? d. If her ejection fraction (EF) is 77%, using the

blood-pressure control each contribute to reducing cardiovascular mortality? 6. Cardiovascular disease is the leading cause of death in the USA. a. What are the two major forms of cardiovascular disease (CVD) and which form is more common in the USA? b. Name three factors that increase the risk of

stroke volume from above, what is her end-diastolic volume (EDV)? e. On average, the normal stroke volume is 70 ml. The HR can drop to as low as 20 bpm when sleeping, while it is normally 70 bpm when

CVD. c. Name a major screening tool for CVD commonly performed during a routine physical. Describe the procedure for the test and specify levels indicative of CVD progression.

awake. Compare the cardiac output when a person is at rest to when they are asleep. f. After a heart attack, a patient’s ejection fraction drops to only 20%. His EDV is 135 ml and heart rate is 90 bpm. Calculate SV and CO and compare these values to the normal values. 4. In our unit on heart failure, we considered what happens to the ejection fraction as disease progresses. What is the ejection fraction? What is a normal value for the ejection fraction? How does the ejection fraction change as heart failure develops? How do we measure the ejection fraction? 5. Cardiovascular diseases were responsible for 871,500 of the 2,428,033 American deaths in 2004. In 2004, death rates from CVD were 335.7 for white males, 448.9 for black males, 239.3 for white females, and 331.6 for black females. (Death rates are given per 100,000 population). From 1994 to 2004, death rates due to CVD declined 25 percent. This decline has been attributed to a combination of smoking cessation, weight loss, and blood-pressure control. a. What percent of all deaths were due to cardiovascular diseases? b. Calculate the number of deaths for each demographic group due to cardiovascular

7. What are the early warning signs of a heart attack? Why is it important to seek treatment quickly in the event of a heart attack? How do these signs differ from heart failure? 8. A patient complains of mild chest pain lasting several minutes at a time and shortness of breath. He says that in just the past week the frequency of these symptoms has increased notably. You order a coronary angiogram. a. Describe the procedure used to obtain this photograph and note any abnormalities. b. Assume that the best films for angiography are acquired if the contrast agent is well mixed in the blood supply. Assume that this occurs after blood re-circulates ten times through the body. The perfusionist tells the X-ray technician that the patient is sedated and has a heart rate of 20 beats per minute. Estimate total blood volume, assume a slightly lower than normal stroke volume of 50 ml, and show a calculation to determine how long the technician should wait before taking the chest X-ray. c. The physician tells the patient that his condition will require treatment and explain that he has three options: coronary artery bypass grafting (CABG), percutaneous transluminal coronary angioplasty (PTCA), or stent implantation. The

Technologies for treatment of heart disease


patient has several questions; please provide answers to each. i. “If autologous tissue is used for the bypass,

a. List three treatment options that can be used in patients with advanced heart failure. b. What are the main advantages and

where do the vessels come from?” ii. “Which procedure will likely require the longest hospital stay?” iii. “Which procedure is least likely to result in restenosis one year after the surgery?” iv. “Which procedure has the highest initial cost?” 9. A 55-year-old man presents to the emergency

disadvantages of each approach? 11. About 4000 people in the United States await heart transplants each year. The first problem all heart transplant recipients face if they survive surgery is donor organ rejection. a. Explain how the immune system rejects the donor heart by considering it a foreign invader. b. Describe the process of organ donor-matching.

department complaining of shortness of breath, swelling in his ankles, tiredness and chest pain. The physician is worried that the patient is suffering from dilated cardiomyopathy, which is a disease that causes the heart to become enlarged. An echocardiogram of the patient’s heart is immediately performed. The physician reviews the results of the echocardiogram and concludes that the maximum volume of blood in the ventricle during a cardiac cycle is 140 ml and the amount of

c. What are the requirements to become an organ donor? Which is the most important step? 12. Heart disease typically develops over a period of decades. a. Describe two approaches to prevent the development of heart disease. b. Has our society focused on prevention of disease or treatment of disease? Should this change? 13. Read the article below and answer the following questions.

blood pumped out of the ventricle each cycle is 60 ml. a. Calculate the end-systolic volume (ESV) and the

March 21, 2004 New Studies Question Value of Opening Arteries

ejection fraction of the patient. Is the ejection fraction normal or abnormal? Explain your reasoning. b. Briefly define systolic failure, diastolic failure

By GINA KOLATA A new and emerging understanding of how heart attacks occur indicates that increasingly popular aggressive treatments may be doing little or

and pulmonary edema. c. Which side of the heart is more likely to be affected first in heart failure? d. Which type of failure do you think that the patient is experiencing? Explain your reasoning. (Hint: take a look at part c.) e. Explain why he is experiencing shortness of breath, tiredness and edema (ankle swelling).

nothing to prevent them. The artery-opening methods, like bypass surgery and stents, the widely used wire cages that hold plaque against an artery wall, can alleviate crushing chest pain. Stents can also rescue someone in the midst of a heart attack by destroying an obstruction and holding the closed artery open. But the new model of heart disease shows that the vast majority of

10. A person dies of cardiovascular disease-related causes every 33 seconds. While the overall rate of heart disease in the USA has shown a promising decline in recent years, one type of heart disease, heart failure, is on the increase. Patients with milder forms can be treated with a number of medications. However, patients with more advanced disease have few options.

heart attacks do not originate with obstructions that narrow arteries. Instead, recent and continuing studies show that a more powerful way to prevent heart attacks in patients at high risk is to adhere rigorously to what can seem like boring old advice – giving up smoking, for example, and taking drugs to get blood pressure under control, drive cholesterol


Biomedical Engineering for Global Health levels down and prevent blood clotting. Researchers estimate that just one of those tactics, lowering cholesterol to what guidelines suggest,

narrowed artery were the culprit, exercise would have caused severe chest pain. Heart patients may have hundreds of vulnerable

can reduce the risk of heart attack by a third but is followed by only 20 percent of heart patients. “It’s amazing and it’s completely backwards in terms of prioritization,” said Dr. David Brown, an interventional cardiologist at Beth Israel Medical Center in New York. Heart experts say they understand why the disconnect occurred: they, too, at first found it hard to believe what research

plaques, so preventing heart attacks means going after all their arteries, not one narrowed section, by attacking the disease itself. That is what happens when patients take drugs to aggressively lower their cholesterol levels, to get their blood pressure under control and to prevent blood clots. Yet, researchers say, old notions persist. “There is just this embedded belief that fixing an artery is

was telling them. For years, they were wedded to the wrong model of heart disease. “There has been a culture in cardiology that the narrowings were the problem and that if you fix them the patient does better,” said Dr. David

a good thing,” said Dr. Eric Topol, an interventional cardiologist at the Cleveland Clinic in Ohio. In particular, Dr. Topol said, more and more people with no symptoms are now getting stents. According to an analysis by Merrill Lynch,

Waters, a cardiologist at the University of

based on sales figures, there will be more than a

California at San Francisco. The old idea was this: Coronary disease is akin to sludge building up in a pipe. Plaque accumulates slowly, over decades, and once it is

million stent operations this year, nearly double the number performed five years ago. Some doctors still adhere to the old model. Others say that they know it no longer holds but that they

there it is pretty much there for good. Every year, the narrowing grows more severe until one day no blood can get through and the patient has a heart

sometimes end up opening blocked arteries anyway, even when patients have no symptoms. Dr. David Hillis, an interventional cardiologist at

attack. Bypass surgery or angioplasty – opening arteries by pushing plaque back with a tiny balloon and then, often, holding it there with a stent – can open up a narrowed artery before it closes

the University of Texas Southwestern Medical Center in Dallas, explained: “If you’re an invasive cardiologist and Joe Smith, the local internist, is sending you patients, and if you tell them they

completely. And so, it was assumed, heart attacks could be averted. But, researchers say, most heart attacks do not occur because an artery is narrowed by plaque. Instead, they say, heart attacks occur when an area of plaque bursts, a clot forms over the area and blood flow is abruptly blocked. In 75 to 80 percent of cases, the plaque that erupts was not

don’t need the procedure, pretty soon Joe Smith doesn’t send patients anymore. Sometimes you can talk yourself into doing it even though in your heart of hearts you don’t think it’s right.” Dr. Topol said a patient typically goes to a cardiologist with a vague complaint like indigestion or shortness of breath, or because a scan of the heart indicated calcium deposits – a

obstructing an artery and would not be stented or bypassed. The dangerous plaque is soft and fragile, produces no symptoms and would not be seen as an obstruction to blood flow. That is why, heart experts say, so many heart attacks are unexpected – a person will be out jogging one day, feeling fine,

sign of atherosclerosis, or buildup of plaque. The cardiologist puts the patient in the cardiac catheterization room, examining the arteries with an angiogram. Since most people who are middle-aged and older have atherosclerosis, the angiogram will more often than not show a

and struck with a heart attack the next. If a

narrowing. Inevitably, the patient gets a stent. “It’s

Technologies for treatment of heart disease


this train where you can’t get off at any station along the way,” Dr. Topol said. “Once you get on the train, you’re getting the stents. Once you get in

arteries where there was too little plaque to be stented or bypassed. Many cardiologists derided him.

the cath lab, it’s pretty likely that something will get done.” One reason for the enthusiastic opening of blocked arteries is that it feels like the right thing to do, Dr. Hillis said. “I think it is ingrained in the American psyche that the worth of medical care is directly related to how aggressive it is,” he said. “Americans want a full-court press.” Dr. Hillis said

Around the same time, Dr. Steven Nissen of the Cleveland Clinic started looking directly at patients’ coronary arteries with a miniature ultrasound camera that he threaded into blood vessels. He found that the arteries were riddled with plaque, but almost none of it was obstructing blood vessels. Soon he began proposing that the problem was not the plaque that produced

he tried to explain the evidence to patients, to little avail. “You end up reaching a level of frustration,” he said. “I think they have talked to someone along the line who convinced them that this procedure will save their life. They are told if you don’t have

narrowings but the hundreds of other areas that were ready to burst. Cardiologists were skeptical. In 1999, Dr. Waters of the University of California got a similar reaction to his study of patients who had been referred for angioplasty,

it done you are, quote, a walking time bomb.”

although they did not have severe symptoms like

Researchers are also finding that plaque, and heart attack risk, can change very quickly – within a month, according to a recent study – by something as simple as intense cholesterol

chest pain. The patients were randomly assigned to angioplasty followed by a doctor’s usual care, or to aggressive cholesterol-lowering drugs but no angioplasty. The patients whose cholesterol was

lowering. “The results are now snowballing,” said Dr. Peter Libby of Harvard Medical School. “The disease is more mutable than we had thought.”

aggressively lowered had fewer heart attacks and fewer hospitalizations for sudden onset of chest pain. The study “caused an uproar,” Dr. Waters

The changing picture of what works to prevent heart attacks, and why, emerged only after years of research that was initially met with disbelief. Early attempts to show that opening a narrowed

said. “We were saying that atherosclerosis is a systemic disease. It occurs throughout all the coronary arteries. If you fix one segment, a year later it will be another segment that pops and gives

artery saves lives or prevents heart attacks were unsuccessful. The only exception was bypass surgery, which was found to extend the lives of some patients with severe illness but not to prevent heart attacks. It is unclear why those patients lived longer; some think the treatment prevented their heart rhythms from going awry, while others say that the detour created by a

you a heart attack, so systemic therapy, with statins or antiplatelet drugs, has the potential to do a lot more.” But, he added, “there is a tradition in cardiology that doesn’t want to hear that.” Even more disquieting, Dr. Topol said, is that stenting can actually cause minor heart attacks in about 4 percent of patients. That can add up to a lot of people suffering heart damage from a

bypass might be giving blood an alternate route when a clot formed somewhere else in the artery. Some early studies indicated what was really happening, but were widely dismissed. As long ago as 1986, Dr. Greg Brown of the University of Washington at Seattle published a paper showing

procedure meant to prevent it. “It has not been a welcome thought,” Dr. Topol said. Stent makers say they do not mislead doctors or patients. Their new stents, coated with drugs to prevent scar tissue from growing back in the immediate area, are increasingly popular among

that heart attacks occurred in areas of coronary

cardiologists, and sales are exploding. But there is


Biomedical Engineering for Global Health not yet any evidence that they change the course of heart disease. “It’s really not about preventing heart attacks per se,” said Paul LaViolette, a senior vice president at Boston Scientific, a stent manufacturer. “The obvious purpose of the procedure is palliation and symptom relief. It’s a quality-of-life gain.” c 2004. From the New York Times, March 21  The New York Times. All rights reserved. Used with permission and protected by the Copyright Laws of the United States. The printing, copying redistribution, or retransmission of the Material without express written permission is prohibited. a. Discuss the advantages of placing more emphasis on prevention of heart disease. b. Discuss the challenges of testing the efficacy of prevention efforts. c. Why do you think our society places so much more emphasis on treatment of end stage heart

disease rather than preventive measures? 14. The heart-lung machine was developed in the 1950s. It enabled many types of cardiac surgery which were previously impossible. a. Describe the function of a heart–lung machine. b. What is the major disadvantage of using the heart–lung machine? c. What are two new technologies which eliminate the need for use of the heart–lung machine? 15. List the major surgical steps involved in implanting an AbioCor total artificial heart.

References [1] World Health Organization. Cardiovascular Disease: Prevention and Control [cited 2007 June 7]; Available from: publications/facts/cvd/en/ [2] Levenson JW, Skerrett PJ, Gaziano JM. Reducing the global burden of cardiovascular disease: the role of risk factors. Preventive Cardiology. 2002 Fall; 5(4): 188–99. [3] Reddy KS, Yusuf S. Emerging epidemic of cardiovascular disease in developing countries. Circulation. 1998 Feb 17; 97(6): 596–601.

[4] Centers for Disease Control and Prevention. Achievements in Public Health, 1900–1999: Decline in deaths from heart disease and stroke – United States, 1900–1999. MMWR. 1999; 48(30): 649–56. [5] Beaglehole R, Irwin A, Prentice T. The World Health Report 2003: Shaping the Future. Geneva: World Health Organization; 2003. [6] Borden WB, Faxon DP. Facilitated percutaneous coronary intervention. Journal of the American College of Cardiology. 2006 Sep 19; 48(6): 1120–8. [7] World Health Organization. Integrated Management of Cardiovascular Risk: Report of a WHO Meeting. Integrated Management of Cardiovascular Risk; 2002 9–12 July. Geneva: WHO; 2002. [8] Merck. Heart Failure. 2005 November [cited 2007 June 7]; Available from: mmpe/sec07/ch074/ch074b.html [9] Waxman S, Ishibashi F, Muller JE. Detection and treatment of vulnerable plaques and vulnerable patients: novel approaches to prevention of coronary events. Circulation. 2006 Nov 28; 114(22): 2390–411. [10] Chandrasekhar A. Patient with Chest Pain. Loyal University Medical Education Network [cited 2007 June 7]; Available from: lumen/meded/mech/cases/case2/Case_f.htm [11] Reul RM. Will drug-eluting stents replace coronary artery bypass surgery? Texas Heart Institute Journal/ from the Texas Heart Institute of St. 2005; 32(3): 323–30. [12] Feder BJ. In the stent era, heart bypasses get a new look. New York Times. 2007 February 25. [13] Sellke FW, DiMaio JM, Caplan LR, Ferguson TB, Gardner TJ, Hiratzka LF, et al. Comparing on-pump and off-pump coronary artery bypass grafting: numerous studies but few conclusions: a scientific statement from the American Heart Association council on cardiovascular surgery and anesthesia in collaboration with the interdisciplinary working group on quality of care and outcomes research. Circulation. 2005 May 31; 111(21): 2858–64. [14] Wimmer-Greinecker G, Deschka H, Aybek T, Mierdl S, Moritz A, Dogan S. Current status of robotically assisted coronary revascularization. American Journal of Surgery. 2004 Oct; 188(4A Suppl.): 76S–82S. [15] Mishra YK, Wasir H, Sharma KK, Mehta Y, Trehan N. Totally endoscopic coronary artery bypass surgery. Asian Cardiovascular & Thoracic Annals. 2006 Dec; 14(6): 447–51. [16] Tung R, Kaul S, Diamond GA, Shah PK. Narrative review: drug-eluting stents for the management of

Technologies for treatment of heart disease restenosis: a critical appraisal of the evidence. Annals of Internal Medicine. 2006 Jun 20; 144(12): 913–19. [17] McMurray JJ, Pfeffer MA. Heart failure. The Lancet. 2005 May 28–Jun 3; 365(9474): 1877–89. [18] Zareba KM. The artificial heart – past, present, and future. Medical Science Monitor. 2002 Mar; 8(3): RA72–7. [19] Gray NA, Jr., Selzman CH. Current status of the total artificial heart. American Heart Journal. 2006 Jul; 152(1): 4–10. [20] Zevitz ME. Heart failure. eMedicine. 2006 June 15; [21] Patwala AY, Wright DJ. Device based treatment of heart failure. Postgraduate Medical Journal. 2005 May; 81(955): 286–91. [22] Groleau R. Operation: Heart Transplant or How to Transplant a Heart in Nineteen Easy Steps. NOVA Online 2000 November [cited 2007 June 7]; Available from: [23] US Department of Health and Human Services. 2007 [cited 2007 June 7]; Available from: [24] Westaby S. The need for artificial hearts. Heart (British Cardiac Society). 1996 Sep; 76(3): 200–6. [25] Song X, Throckmorton AL, Untaroiu A, Patel S, Allaire PE, Wood HG, et al. Axial flow blood pumps. Asaio J. 2003 Jul–Aug; 49(4): 355–64.


[26] Poirier VL. The LVAD: A case study. The Bridge. 1997 Winter; 24(4). [27] Artificial heart: the debate goes on. Science News. 1986 February 22; 129: 122. [28] Bourque K, Gernes DB, Loree HM, Richardson JS, Poirier VL, Barletta N, et al. HeartMate III: pump design for a centrifugal LVAD with a magnetically levitated rotor. Asaio Journal. 2001; 47: 401–5. [29] Fukamachi K. New technologies for mechanical circulatory support: current status and future prospects of CorAide and MagScrew technologies. Journal of Artificial Organs. 2004; 7(2): 45–57. [30] Stevenson LW, Rose EA. Left ventricular assist devices: bridges to transplantation, recovery, and destination for whom? Circulation. 2003 Dec 23; 108(25): 3059–63. [31] Chen JM, Naka Y, Rose EA. The future of left ventricular assist device therapy in adults. Nature Clinical Practice. 2006 Jul; 3(7): 346–7. [32] Mitka M. Heart disease a global health threat. Jama. 2004 Jun 2; 291(21): 2533. [33] Brown AI, Garber AM. A concise review of the cost-effectiveness of coronary heart disease prevention. The Medical Clinics of North America. 2000 Jan; 84(1): 279–97, xi. [34] Nova Transcripts, “Electric Heart”, PBS Airdate December 21, 1999, transcripts/2617eheart.html.

13 Clinical trial design and sample size calculation

In Chapter 12, we considered the many limitations of the currently available treatments for heart disease; because these invasive procedures are expensive and have serious side effects, there is an important global need to develop more cost effective ways to prevent heart disease. In the early 1990s, a series of small, highly publicized studies offered hope that a simple intervention – taking high doses of vitamin E (Figure 13.1) – might reduce the risk of developing cardiovascular disease by as much as 40% [1, 2]. A subsequent randomized clinical trial compared the rates of myocardial infarction and cardiovascular death in a group of 1035 patients taking vitamin E to those in a group of 967 patients taking a placebo; results were reported in 1996 and also showed that vitamin E provided a protective effect [3]. Yet, a pivotal clinical trial involving 9541 patients in 2000 indicated that there might be no reduction in the risk of cardiovascular disease for those who take vitamin E supplements and when these patients were followed for more than seven years, it appeared that the use of vitamin E supplements may actually increase the risk of developing heart failure [4, 5]. This series of studies provides a good illustration of both the process and challenges of clinical research. Generally, early studies involving small numbers of patients provide data which allow us to generate

Figure 13.1. There has been considerable disagreement among different epidemiologic and clinical studies designed to determine whether taking vitamin E supplements can reduce the risk of developing cardiovascular disease.

hypotheses; these hypotheses must then be tested rigorously in larger clinical studies. Because of inherent biological variability, it is not uncommon for results of early studies to be contradicted by larger, more rigorously designed studies. In fact, a recent study compared conclusions presented in highly cited articles to those of subsequent studies with larger sample size or better controlled design [6]. Results showed that nearly 1/3 of highly cited studies were later contradicted and that

Clinical trial design and sample size calculation


this was most likely for studies where patients were not randomly assigned to a treatment or control group! In this chapter, we will examine how clinical studies

As we consider these important issues, we will find that the process of designing clinical research studies and clinical trials has many similarities to the scien-

and clinical trials are designed. Our goal is to develop the tools to properly interpret and use the results of clinical research. We begin by examining how to characterize biological data and its associated variability. Next, we provide an overview of the different types of clinical studies and clinical trials used to generate and test hypotheses. Finally, we examine how to calculate whether a study to test a hypothesis includes sufficient

tific method that we learned about in Chapter 7. In order to help form hypotheses, we begin by carrying out observational and epidemiologic studies. Ultimately, we must carry out carefully controlled, randomized clinical trials designed to isolate the effect of just one factor and determine whether it has the predicted impact.

numbers of patients to ensure that any resulting conclusions are based on real differences in the data and not just statistical variability.

The challenge in this process is that because of biological variability, we must always ensure that studies include a large enough group of patients to be sure that any differences we see are real and not just due to chance.

Descriptive statistics We can use descriptive statistics to analyze the average core body temperature for normal adults. The following dataset contains the core body temperature from a group of 65 men and 65 women. We can calculate parameters which measure the central tendency of the data as well as the dispersion in the data [7]. Using the definitions in the text, we find the following values. Mean core temperature: 98.25 ◦ F Median core temperature: 98.3 ◦ F Mode core temperature: 98 ◦ F Standard deviation: 0.73 ◦ F Using a bin size of 0.4 ◦ F, starting at a minimum value of 96.3 and ending at a maximum value of 100.4, we find the frequency histogram shown below (Journal of Statistics Education, Volume 4, Number 2 (July 1996)).


Biomedical Engineering for Global Health

Descriptive statistics In clinical research, we are constantly concerned with determining whether differences observed between groups of patients are due to an intervention given only to one group or are simply due to biological variability. Generally speaking, we carry out clinical trials in groups of patients that we believe are representative of the greater population. How do we characterize these groups? An important first step is to use descriptive statistics to assess demographic variables (e.g. gender, age) for the patients in our sample and to assess clinical parameters for these patients (e.g. blood pressure, serum nicotine levels). When characterizing data like this, we generally need to assess three main factors: the central tendency of the data (e.g. mean, mode), the spread in the data (e.g. variance, standard deviation), and how the data are distributed across the range of possible values (e.g. normally distributed) [8]. There are several methods to characterize central tendency of the data. If we are interested in a single, continuous variable, x, we define the mean to be the average of measured values for our population [8]. The median is that value which occupies the middle rank when the values are rank ordered from least to greatest. The mode

Figure 13.2. Normal distribution. Gore and Altman, BMA London. Copyright BMJ Publishing Group, with permission.

tributions in biological data is the normal distribution (Figure 13.2). In normally distributed data, the relative frequency histogram is determined completely by the mean and the standard deviation, and the mean, mode and median are identical. One standard deviation on either side of the mean contains 68% of the area under the relative frequency histogram for a normal distribution: 95% of the data are contained within +/−1.96 SD

is the most commonly observed value. Similarly, there are several measures of the dispersion of the data. Two of the most common are the variance and the standard deviation. The variance is the sum of

of the mean.

the squared deviations from the mean.

Types of clinical studies (13.1)

There are two major types of clinical studies: those conducted with the goal of hypothesis generation and those designed to test a specific hypothesis. Hypothesis

The standard deviation is the square root of the variance.

generating studies are often referred to as epidemiologic studies, while hypothesis testing studies are called clinical trials.

Variance =


¯ (Xi − X)



In order to examine the distribution of our data, we can display the data graphically in a histogram, in which we divide the data into bins and count the number of values that fall into each bin. A relative frequency histogram is a plot of the fraction of the observations that fall within that bin. The area under a relative frequency histogram is unity. The distributions of data often tend to follow similar curves. One of the most commonly encountered dis-

Hypothesis generation There are several types of studies which can be used to generate clinical hypotheses. For example, in a case study, we identify one patient and attempt to understand the factors related to their illness, either by reviewing their history or through physical examination or clinical testing. Similarly, in a case series, we identify a group

Clinical trial design and sample size calculation


Infant Diagnosis: June 13, 2007 Kim Malawi The project that I’m going to be spending the rest of my summer working on is a Malawi Ministry of Health pilot project that is supported by a variety of NGOs, including BIPAI (Baylor International Pediatric AIDS Initiative), UNICEF, the Clinton Foundation, etc. It is a part of their “HIV-Free Generation” goal. This project aims to identify HIV exposure or infection as early as possible to begin providing care as early as possible. Ideally, all pregnant women will be routinely screened for HIV at their prenatal care visits (ANC). Those that are pregnant will get CD4 counts, be clinically staged and given a single dose of Nevirapine (NVP) to take at the onset of labor. If they qualify (by CD4 count or WHO stage) they can be referred to an ARV clinic and started on ARVs. The single dose of NVP isn’t needed for women on full ARV therapy, but does reduce mother to child transmission by about 50% if the women aren’t otherwise on therapy. (The baby is also supposed to receive a dose when it is born for this to be most effective.) Since this is obviously not ideal, right now the program is also involved in identifying infants below the age of 18 months, which is an age group that cannot be identified by the currently available rapid HIV tests. Those tests check for anti-HIV antibodies in the blood, but babies can have their mother’s antibodies (passed through the placenta) for up to 18 months. But, we are now able to do DNA PCR testing on dry blood spots. These tests identify viral DNA in the blood, which you only have if you are infected. It can identify babies as early as six weeks old. The only problem is that while a positive is a definite positive, a negative is not a definite negative in infants that age. Continued breastfeeding is continued exposure, so infants have to be retested six weeks after they stop breastfeeding. And breastfeeding is a huge problem. It would be best to not do it at all, but babies would almost all starve to death. Formula is far too expensive. So most women breastfeed. In that situation, it would be best to breastfeed exclusively for six months and then rapidly wean to solid foods/juice. But they also give other foods (mixed feeding). The problem with mixed feeding is that it actually increases the chances of transmission through a couple mechanisms. First, anything other than breast milk acts as an irritant and makes micro perforations in the baby’s intestines, essentially breaking a barrier between the baby’s body and maternal HIV. Secondly, the baby’s body recognizes anything but mother’s milk as foreign, which initiates an immune response, drawing CD4 cells to the intestinal tract, and thereby bringing the HIV’s host cells into close proximity to the virus. Many mothers start mixed feeding as young as two months old because they are culturally pressured to do so (if you don’t give your baby other food as young as possible, you clearly don’t love them enough) and continue with the mixed feeding until nearly two years old. Right now our job is to train health care providers at these seven pilot sites about the new testing algorithm and procedures. We then provide a week to two weeks of onsite support and coaching to make sure they get it right. We are also attempting to identify potential HIV+ kids in their pediatric wards, under five clinics and outpatient clinics. So, that’s what I’m up to. It’s very interesting and very exhausting.


Biomedical Engineering for Global Health

of patients with a similar illness and try to ascertain similar factors that may be responsible for their disease. We have already seen a good example of a case series. Recall that, in 1981, a series of four previously healthy homosexual men were identified who developed Pneumocystis carinii pneumonia and mucosal candidiasis – the first description of AIDS and its association with behavioral risk factors.

Hypothesis testing Once we have formulated a hypothesis, there are a number of different types of clinical studies which can be used to test the hypothesis. Hypothesis testing studies

How to find a clinical trial The National Institutes of Health maintains a publicly accessible database of clinical trials to test experimental therapies for serious or life-threatening diseases conducted with federal and private support. The database provides a summary of the purpose of the study, patient recruiting status, criteria for patient participation, the location of the trial, the

are divided into two major groups: observational and experimental. In an observational study, we identify a

research study design, the phase of the trial, and the disease or condition and drug or therapy under study. Currently, information can be found for more than 36,000 ongoing trials [10].

group of patients with disease and a similar group of patients without disease. We collect data from these two groups of patients and compare them to test our hypothesis. Observational studies have the advantage

first planned, then data are collected; this enables the investigators to manipulate the intervention in a con-

of being relatively simple and inexpensive to carry out, but they can suffer from important problems. One of the most common is bias – suppose we identify a group of patients with lung cancer and a matched group of healthy controls. We notice that the lung cancer patients report an increased frequency of consuming alcohol, and we conclude that alcohol consumption may be related to the development of lung cancer. It is also possible that alcohol consumption is strongly correlated with cigarette smoking and that smoking is actually the factor which is responsible for the increase in lung cancer. In an observational study, we can’t control the administration of the intervention to decisively show cause and effect – we can only retrospectively analyze differences in our two groups of patients [9]. To avoid this and other sources of error, we can design experimental approaches to test our hypothesis. A clinical trial is a research study to evaluate the effect of a new diagnostic or therapeutic in a group of patients [11]. Clinical trials help us to determine whether new interventions are safe and effective. Clinical trials may involve a single group of patients or multiple groups of patients, but in general, their goal is to isolate all but a single variable and measure the effect of that variable. Clinical trials are done in a prospective way: they are

trolled setting to show cause and effect [9]. A single arm study involves a single group of subjects; subjects receive an intervention and then we monitor them to see if their condition improves [12]. Usually, we compare the change in their condition before and after the intervention (change from baseline) so that each patient serves as their own control. One problem with a single arm study is that improvements following an intervention may be partially due to the placebo effect – where patients or their physicians think they are getting better simply because they believe they are receiving an intervention – and this can change their perception of symptoms. To avoid this problem, clinical trials can be carried out using two groups of patients, one of which receives the intervention and the other which receives a placebo. Randomized clinical trials, in which different subjects are randomly assigned to different treatment and control groups, are generally considered to be the strongest type of clinical evidence [12]. In a doubleblinded, randomized clinical trial, participants are randomized in a way so that neither the participants nor their physician knows which group they are part of. This type of study design eliminates conscious bias. Subjects must be randomized in a way so that the treatment and

Clinical trial design and sample size calculation


control groups are similar on average; in this way we can attribute any differences in the outcomes between the groups to the intervention and not any other factors that may have differed between the groups.

Clinical trials: drugs Clinical trials of new drugs are carried out in different phases to assess the safety and efficacy of the new drug. In a phase I trial, researchers assess the safety of the drug in a series of 20–80 normal volunteers. In a phase II trial, the experimental drug is given to a larger number of volunteers (100–300) to assess both its safety and efficacy. In a phase III trial, the new drug is given to a large number of patients in order to compare its effectiveness to standard therapies and to monitor for side effects. Typically, phase III trials are randomized, placebo controlled, double-blinded studies. It is especially important to carry out a rigorous calculation of the sample size required for phase III trials.

Sample size The sample size is the number of subjects in each arm of a study needed to detect a predetermined difference [13]. Determining the necessary sample size for a clinical trial is a careful balancing act – as the sample size becomes larger, we reduce the risk that differences between the two groups arise due to chance, but the cost and complexity of doing the study increase. Our goal in setting a sample size for a randomized clinical trial is to appropriately balance these factors. In general, setting a sample size is a complex statistical calculation. However, for data which can be described by the normal distribution, there is a relatively straightforward process to estimate the number of patients required. Let’s examine the process to determine the sample size for a two-arm, randomized clinical trial. In this type of trial, we have both a treatment group (receives the intervention) and a control group (receives a placebo). We will monitor the clinical outcomes

Figure 13.3. To design a clinical trial, you must be able to fill out all the information in the template shown.

in each group. We must select a primary outcome – this is the outcome we are most interested in comparing between the two groups. In selecting our sample size, we want to ensure that any differences between the treatment and control group are real, knowing that there will be some statistical uncertainty associated with the primary outcome. We will choose our sample size so that this uncertainty is sufficiently lower than the difference in primary outcome we wish to detect between the control and treatment groups. Figure 13.3 provides a template which illustrates all of the information needed to calculate the sample size for such a trial. What happens if we choose our sample size to be too small? Essentially, there are two types of mistakes that can result. The first is that we mistakenly conclude that there is a difference between the two groups, when in reality there is no difference. This is called a Type I error, and can result in adopting an intervention, even when it is not truly effective. The second type of error that we can make is to mistakenly conclude that there is not a difference between the two, when in reality there is a difference. This is called a Type II error and can result in us discarding a potentially effective intervention. Our sample size will be dictated by the risk that we are willing to accept for making each of these types of errors. The risk of making a Type I error is called the significance level of the study. Typically, studies are designed so that the risk of making a Type 1 error (significance level) is between 1% and 5%. The risk of making a


Biomedical Engineering for Global Health

Type II error is usually denoted by the variable β. Typically, studies are designed so that the risk of making a Type II error is between 10% and 20%. Often, we refer to the power of a study, where the power is defined to be 1 – the probability of making a Type II error. The power of a study usually ranges from 0.8 to 0.9. With these definitions, we can now outline the process of calculating a sample size [14]. First we must select a primary outcome. Our primary outcome can be either a binary measure (did the patient have a heart attack?) or a continuous variable (what is the patient’s systolic blood pressure?). If the outcome variable is binary, we must estimate the expected rate of the primary outcome in the treatment group and the control group. If the primary outcome variable is continuous, we must estimate the size of the difference in the average outcome for the two groups and the expected standard deviation associated with this variable. Next we set acceptable levels of Type I and II errors. This determines the p-value and the power of the study. With this information, we can use a graphical tool called Altman’s nomogram to estimate the sample size required in each group. To use Altman’s nomogram, we must first calculate the standardized difference we expect to find in the primary outcome variable. For a binary outcome variable, the standardized difference is calculated as: P1 − P2 standardized difference =  p¯ (1 − p¯ ) p¯ =

P1 + P2 , 2



where P1 is the fraction of patients in the treatment group who experience the primary outcome and P2 is the fraction of patients in the control group who experience the primary outcome. For a continuous outcome variable, the standardized difference is calculated as: effect size , standardized difference = standard deviation


where the effect size is the difference in the average value of the primary outcome we expect between the treatment and control groups and the standard deviation is that expected for the primary outcome variable.

Figure 13.4. Altman’s nomogram, showing sample size calculation for a standard difference of .85 and a power of .8. Altman’s nomogram is used to calculate the sample size required in each arm of a study, given the expected standardized difference, desired power and significance level. Note that for the results obtained using Altman’s nomogram to be valid, we assume that the continuous variable follows the normal approximation to the binomial distribution and the two groups being compared are of equal size. Altman (1982). How large a sample? In Statistics in Practice. Eds S. M. Gore and D. G. Altman. BMA London. Copyright BMJ Publishing Group, with permission.

We use Altman’s nomogram (Figure 13.4) to determine the number of subjects required in each group by drawing a straight line to connect the standardized difference we wish to detect and the desired power of the study. The point at which this line intersects the line drawn at the desired significance level gives the sample size required for each group. Suppose we wish to carry out a randomized clinical trial to compare the effectiveness of a new drug eluting stent to that of the standard uncoated stent. We design a study where patients requiring treatment for coronary artery disease are randomized to receive either the standard stent or the new drug eluting stent. We decide that our primary outcome will be to follow the patients and determine whether or not they experience restenosis

Clinical trial design and sample size calculation during the first six months following the procedure. This is an example of a binary outcome variable, since each patient either does or does not experience restenosis. A pilot clinical trial with the new drug eluting stent suggests that only 10% of patients experience restenosis following treatment with this device. A review of the literature shows that, on average, the number of patients who experience restenosis within six months following treatment with a standard stent is approximately 45%. We wish to design a trial with a 5% significance level and a power of 0.8, meaning that we are willing to take a 5% chance of making a Type I error and a 20% chance of making a Type II error. We begin by calculating the standardized difference: P1 − P2 standardized difference =  p¯ (1 − p¯ ) 0.45 − 0.1 = 0.78 = √ 0.275(1 − 0.275) (13.5)


P1 + P2 0.45 + 0.1 = = 0.275. 2 2


With this, we turn to Altman’s nomogram, and draw a line connecting a power of 0.8 to a standardized difference of 0.78. We look to see where this line intersects that representing a 0.05 significance level and see that 54 patients will be required in each arm of the study. In the case of a continuous outcome variable, we follow an identical procedure, except when calculating the standardized difference. In summary, when designing and analyzing clinical trials there are four major parameters of interest. 1. 2. 3. 4.

The significance level of the study. The power of the study. The effect size one can detect. The sample size.

For a fixed significance level and power, as we increase the sample size we can detect smaller and smaller differences in the outcomes between the treatment and control groups. It is important to place the effect size we wish to determine in the proper clinical context. There is an important difference between the statistically significant difference we can detect given the


parameters above and what may be clinically meaningful. The smallest difference that is clinically meaningful is sometimes called the minimum clinically important difference (MCD). It is important not to waste time and money enrolling patients in order to detect differences that are substantially less than the MCD, since these differences are of no clinical importance. We have seen that in order to calculate a sample size, we need to estimate the difference that might exist between the treatment and control groups and the expected variance in the data. In practice, the difference may be larger than predicted or the spread in the data may be smaller than expected. In this case, the number of patients required to reach statistical significance will be smaller than initially predicted [13]. However, it is also possible that the differences between the treatment and control groups will be smaller than predicted or the spread in the data will be larger; in order to reach significance, more patients will be required than originally estimated [13]. Frequently, special committees called Data Safety and Monitoring Boards (DSMB) are convened in order to monitor interim data in clinical trials to ensure the safety of participating subjects [15]. The federal government requires that a DSMB be used to monitor all phase III clinical trials. Typically, a DSMB is composed of scientists and clinicians who are knowledgeable about the field of interest, but are not otherwise involved in the study. The role of the DSMB is to analyze all adverse events reported in a trial and to perform interim analysis of clinical outcomes to determine whether a trial should be stopped early. If an interim analysis indicates that a new treatment is substantially better than the standard of care, then the study may be stopped early and the new treatment offered to both the study group and the control group [11]. Conversely, if an interim analysis reveals substantially increased risk associated with a new treatment, then a study may also be stopped early to prevent additional harm. Making decisions to stop a trial early are frequently difficult, because they force researchers to make judgments about when interim results cross the boundary from suggestive to conclusive. Making decisions to stop a trial or proceed with a trial following an interim analysis can also raise serious ethical questions.


Biomedical Engineering for Global Health

Weekend in Mochudi and watching the Zebras (soccer team): June 18, 2007 Rachel Botswana After a PAC doctor, Heather’s suggestion and Liz’s approval, we went ahead and started screening patients in hopes of casting a larger net to catch the patients with less then perfect adherence (although for the most part, all the patients in Gabarone have good adherence, especially when compared to Francistown, a clinic where the pharmacist constructed a list of 31 reasons why he would not do pill counts, and is described as eccentric . . . ) which brings me to an idea for another project at Rice which would not be life changing here, but would be a pleasant convenience – a pill counter! I am thinking something akin to those change counters where you dump the change on top and it sorts it and counts it. The most important part would be cost, simplicity and ease of use so that at the Francistown clinic, there would be no excuse for not pill counting, a practice which reminds me of DOTS (directly observed therapy) for TB. While not being exactly direct, DOTS is helpful for making sure people continue to take their pills and be counseled when they miss doses. After talking to the nurse head, Mamapula (which means Mother of Rain) I found out that we probably should’ve been communicating better. She said she was a little in the dark as to what exactly we were doing, but after about 5 minutes explaining, she was good to go and helped show us how the system works at reception – where to pick up the binders of the patients to be seen – and how to do the pill counts – where to find the pharmacy sheet of how many pills the patient left with. I guess we hoped to expedite the process in the exam rooms with the doctors, but we weren’t able to “catch” all the patients that we had hoped to, only ending up with a couple. On the other hand, considering that the patients are taking their pills so well, this is great news for the clinic! Tomorrow we will go back to the data extraction project in an effort to see the trends in the resistance tests (RT) that patients who were failing on line 1 or 2 of the ARVs. RTs are expensive and usually only done in extreme circumstances but the results of collecting the mutations of the HIV are pivotal in letting the doctor know which drugs will be ineffective and which ones will work.

For example, a recent trial was carried out to assess the efficacy of a new treatment for severe sepsis (bloodstream infection) [15]. Researchers designed a trial to compare a new drug, tifacogin, to that of a placebo; the primary outcome was mortality. The trial was designed

group and 29.1% in the tifacogin group. This difference was significant at the 0.006 level. The DSMB was forced to decide whether to stop the trial early, given the apparently lower mortality rate associated with the new drug. Ultimately, the board decided that the results were

with a sample size of approximately 1500 subjects. An interim analysis conducted after 722 patients had been accrued showed a mortality rate of 38.9% in the placebo

not sufficiently compelling and the trial was continued. Neither the researchers nor the prospective trial participants were informed of the interim results.

Clinical trial design and sample size calculation


Clinic: dosing guide: June 22, 2007 Lindsay Botswana As we still attempt to fashion some sort of trial for our dosing guide, we are working on another project Dr. Lowenthal presented to us. For our “trial,” we just don’t have enough patients with poor adherence to create any kind of pool big enough to make our results somewhat reliable. For now, we’re for the most part just giving everybody dosing guides and not creating a control group (those w/o guides). As unscientific as this is, the doctors and nurses are referring patients to us whom they think really need the guide, at a rate of one to two per day. Given the current situation, it makes more sense to just allow the people who need guides to get them – we won’t be able to bring together a “scientific” study. There are many confounding factors in creating a “scientific” study, mainly that all patients referred to us will also have to go through adherence class again, which itself may raise adherence rates. I would rather rely on the patients’ comments when they come back to the clinic for their next appointment to see if the guide is helpful. I think this may end up being a lot more subjective than we all expected, but I’m okay with that. We had an 11-year-old patient who came in the other day with her mother, who had recently taken custody of her due to the death first of a nanny and then of an aunt, who had been caring for the girl very well and making sure she took her meds. After I created a dosing guide for the mother (the girl wasn’t with us), she went back to the waiting room as usual. As I went back to the nurses to make sure the guide was correct, they told me about the family’s situation – the mother was quite irresponsible, and the 11-yr-old girl was taking the ARVs on her own! She had been doing a very good job despite the complexity of her regimen, but she still needed help, which she wouldn’t get from her mother. Her eyes lit up as I explained to her the dosing guide and how she could use it, as the mom’s eyes wandered around the room in boredom. I think we are better off helping the five or six children needing help who come in each week, rather than spending our time creating a trial. A doctor at clinic suggested that we try to set up a visit at one of the SOS Villages (orphanages) to see if we would be able to give all the “mothers” there dosing guides for all the kids who need them.

The importance of the minimum clinically important difference “Samples which are too small can prove nothing; samples which are large enough can prove anything” [14].

Was this decision scientifically sound? Was it ethical? The interim analysis, although not definitive, suggested that the new drug reduced mortality by about 25% compared to placebo. If the researchers had been informed of the interim results, it might have made them more hesitant to enroll their patients in the trial and to fol-

low the study protocol. In the end, recruitment was continued to the final sample size. When all the data were analyzed, it was determined that mortality in the tifacogin group was 34.2% compared to 33.9% in the placebo group – a difference that was neither clinically nor statistically significant. To address these ethical concerns regarding decisions made by DSMBs, it has been recommended that informed consent documents be modified to indicate if a trial will be monitored by a DSMB. If so, during informed consent, the role of the DSMB should be explained, noting that the DSMB may make recommendations to continue a trial even in the face of evidence suggesting


Biomedical Engineering for Global Health

effectiveness of one of the treatments. The informed consent process should make clear that, in the interests of maintaining the scientific integrity of the study, interim results will not necessarily be made available to patients enrolled in the study [15].

Bioengineering and Global Health Project Project task 9: Design a clinical trial to test your new technology In this task, you will design a clinical trial to test the new technology that you have developed. As part of this design, you must choose the subjects to be tested including the control group (receives standard of care) and the experimental group (receives new technology). You must specify the primary and secondary outcomes that you will monitor in the trial. You must calculate a statistically justified sample size for the trial. Complete the table below and provide a one page summary of the trial design and sample size justification.

Print a certificate of completion upon completing the course. 2. You are a researcher working on a vaccine for malaria, which is caused by a parasitic protist. You are interested in performing clinical trials for your vaccine. a. During phase I testing, you must recruit about 100 healthy volunteers. What are you trying to determine through this testing? b. During phase II testing, what should your malaria vaccine be tested against? c. Phase III testing involves a double-blind study. What does this mean and why is it important? 3. You are designing a study to test a new implantable artificial kidney for patients with end-stage renal disease. You divide your patients into two groups: one will receive an implanted artificial kidney, the other will receive tri-weekly hemodialysis (standard of care in end stage renal disease). Your primary endpoint is mortality for all causes at one year, a secondary endpoint is patient quality of life at one year, which will be assessed via questionnaire. a. What is a Type I error? What are the possible consequences of making a Type I error in this study? b. What is a Type II error? What are the possible consequences of making a Type II error in this study? c. Define the p-value. d. Define power (if you use Greek letters in your definition, you must define these as well). e. Why would blinding be difficult in this study?

Homework 1. Complete the Human Participant Protections Education for Research Teams course which can be found at: learning/humanparticipant-protections.asp. This course presents information about the rights and welfare of human participants in research. The tutorial is designed for those conducting research involving human participants. Complete the exercises at the end of each of the six content areas.

f. Which of your endpoints is more likely to be affected by a lack of blinding? g. Assuming you expect 30% mortality at one year for the control group and 20% for the treatment group, what sample size would be required to achieve 80% power? What p-value should you use? Justify this value. 4. You are designing a clinical trial to compare the performance of a new thrombolytic agent to dissolve blood clots associated with acute myocardial infarction. You design a trial to compare the new agent to the standard of care, streptokinase. You choose the 30 day mortality as your primary

Clinical trial design and sample size calculation endpoint. There will be some statistical uncertainty associated with the measured mortality rate in the treatment and control groups. Your goal in selecting the sample size for the trial is that this uncertainty be significantly less than the difference in the mortality rate between the control and treatment groups. We must set acceptable levels for the risks of Type I and II error. a. Define Type I error and Type II error. b. Suppose you expect a mortality rate in the group treated with the new drug stent of 5%, while the expected restenosis rate in the group treated with the current stent is 7%. You calculate a standardized difference of 0.19. If you can tolerate a 20% risk of Type II error and a 5% risk of Type I error, how many patients are needed in the trial? Use Altman’s nomogram to indicate how you calculated your answer. c. If the mortality rate for the new drug was expected to be 1%, would the required sample size increase or decrease? d. List one secondary outcome you would want to monitor in this trial. 5. Consider the Acute Respiratory Distress Syndrome Network trial of low versus traditional tidal volume ventilation in patients with acute lung injury and acute respiratory distress syndrome published in 1996. Mortality rates in the low and traditional volume groups were 31% and 40%, respectively, corresponding to a reduction of 9% in the low volume group. What sample size would be required to detect this difference with 90% power using a cut-off for statistical significance of 0.05? Source: [16].

References [1] Rimm EB, Stampfer MJ, Ascherio A, Giovannucci E, Colditz GA, Willett WC. Vitamin E consumption and the risk of coronary heart disease in men. The New England Journal of Medicine. 1993 May 20; 328(20): 1450–6. [2] Stampfer MJ, Hennekens CH, Manson JE, Colditz GA, Rosner B, Willett WC. Vitamin E consumption and the risk of coronary disease in women. The New England Journal of Medicine. 1993 May 20; 328(20): 1444–9.


[3] Stephens NG, Parsons A, Schofield PM, Kelly F, Cheeseman K, Mitchinson MJ. Randomised controlled trial of vitamin E in patients with coronary disease: Cambridge Heart Antioxidant Study (CHAOS). The Lancet. 1996 Mar 23; 347(9004): 781–6. [4] Yusuf S, Dagenais G, Pogue J, Bosch J, Sleight P. Vitamin E supplementation and cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. The New England Journal of Medicine. 2000 Jan 20; 342(3): 154–60. [5] Lonn E, Bosch J, Yusuf S, Sheridan P, Pogue J, Arnold JM, et al. Effects of long-term vitamin E supplementation on cardiovascular events and cancer: a randomized controlled trial. Jama. 2005 Mar 16; 293(11): 1338–47. [6] Ioannidis JP. Contradicted and initially stronger effects in highly cited clinical research. Jama. 2005 Jul 13; 294(2): 218–28. [7] Shoemaker AL, College C. What’s normal? – temperature, gender, and heart rate. Journal of Statistics Education. 1996; 4(2). [8] Neely JG, Stewart MG, Hartman JM, Forsen JW, Jr., Wallace MS. Tutorials in clinical research, part VI: descriptive statistics. The Laryngoscope. 2002 Jul; 112(7 Pt 1): 1249–55. [9] Bernson M. Tutorial: an introduction to clinical trials Part I of II: purposes and phases. The Next Generation. 2005; 1(4). [10] US National Institutes of Health. 2007 [cited 2007 June 7]; Available from: http://www. [11] The Patient Education Institute. Clinical Trials. MedlinePlus 2005 January 25 [cited 2007 June 7]; Available from: medlineplus/tutorial.html [12] Bernson M. Tutorial: An introduction to clinical trials part II of II: statistics and experimental design. The Next Generation. 2005; 1(5). [13] Neely JG, Karni RJ, Engel SH, Fraley PL, Nussenbaum B, Paniello RC. Practical guides to understanding sample size and minimal clinically important difference (MCID). Otolaryngol Head Neck Surgery. 2007 Jan; 136(1): 14–18. [14] Ogundipe L. Sample size determination in clinical research: 2. Hospital Medicine. 2000 Nov; 61(11): 797–8. [15] Slutsky AS, Lavery JV. Data safety and monitoring boards. The New England Journal of Medicine. 2004 Mar 11; 350(11): 1143–7. [16] Whitley E, Ball J. Statistics Review 4: Sample size calculations. Critical Care. 2002 May 10; 6: 335–341.

14 Technology diffusion

We have considered the development of new technologies to improve the prevention of infectious disease, the early detection of cancer and the treatment of cardiovascular disease. Through this journey, we have seen the often integrated processes of scientific research, engineering design and clinical research in action. Through this coordinated process, we develop new healthcare technologies which are safe and effective. How do we ensure that the results of these efforts actually reach the patients who need them? We will now consider this important question, as we examine how new health technologies are managed. In the present chapter we will examine the diffusion of new innovations – what drives the adoption of technologies that are newly proven to be beneficial? How does the diffusion of new technologies differ throughout the world? How can we speed the adoption of effective new technologies? The diffusion of new interventions has been historically slow. Consider the example of the use of vitamin C to treat scurvy [1]. Scurvy was a major cause of mortality for early sailors. In 1497, Vasco Da Gama lost 100 out of 160 crew members to scurvy sailing around the Cape of Good Hope. A dietary connection was suspected, and in 1601, British Navy Captain James Lancaster was in command of four ships traveling from England to India when he performed a clinical trial to test whether lemon juice could reduce the incidence of scurvy. He required

Figure 14.1. Lemon juice, a simple intervention to prevent scurvy, was not adopted until more than 200 years after first proven effective.

sailors to take three teaspoons of lemon juice daily on one ship, while sailors on the other three ships served as the control group. The results were astonishing; while 110 of 278 (40%) sailors died in the control group, there were no deaths in the experimental group. Despite this impressive result, the intervention was not widely adopted. In fact, in 1747, British Navy Physician James Lind repeated the study with similar results. It was not until 1865 that the British Navy finally adopted the innovation, 264 years after Captain Lancaster’s successful clinical trial!

Technology diffusion


It is interesting to think about how the people responsible for innovations and their adoption interact during this process. The group of people who most rapidly adopt new technologies are the innovators who develop these new technologies. Innovators are willing to take risks to develop new technologies; they are often perceived as risk takers or mavericks, who may be overly invested in the success of their own innovation. As a result, they may be viewed as incautious and are frequently socially disconnected. Next to adopt a new technology are the early adopters; this group of people is usually involved in testing and comparing the efficacy of several new interventions in order to determine which is most effective. Early adopters are viewed as well connected, social opinion leaders; their actions are watched by the communities in which they practice. The early

Figure 14.2. Innovation adoption curve [3]. Courtesy of

majority generally follows the lead of early adopters to accept new innovations. They rely on the conclusions of people they know before deciding to test a new change. The late majority wait to see local proof that a new inter-

Has the situation changed today? Remarkably, a recent study found that it now takes an average of

vention is effective before adopting it. Finally, the most conservative group, sometimes called laggards, wait to adopt a new innovation until it has become the status

17 years for new knowledge generated in randomized clinical trials in the United States to be implemented in clinical practice, and even then, implementation is highly non-uniform [2]. Technology diffusion is even

quo or standard of care. Let’s examine a case study of the diffusion of a relatively new technology – that of laparoscopic cholecystectomy – or removal of the gall bladder using

slower in developing countries, which frequently do not have the capacity to carry out medical research, and

fiber optic guided surgery. This procedure was one of the first minimally invasive surgical procedures to be

must instead import new technology developed in other settings [3]. Let’s examine the pace at which adoption of new technologies proceeds. If we graph the fraction of a target population who has adopted a new technology over time, we find that the curve frequently has an S-shape

developed and the story of its adoption illustrates many of the concepts we have just introduced. The gall bladder (Figure 14.3) acts to store bile made by the liver. After you eat a meal, the gall bladder contracts, and secretes bile into a duct which empties into the small intestine. Bile provides an important aid in

(Figure 14.2). Initially, diffusion is slow. As more of the population adopts the innovation, the speed of diffusion increases. Often, we find that there is a tipping point, when between 15% and 20% of the population have

digestion. In some cases, liquid bile may precipitate into solid stones called gall stones. This is a relatively common occurrence, with nearly 1/5 of North Americans and 1/4 of Europeans developing gall stones at some

adopted an innovation, its spread becomes difficult to stop. At some point, the market is nearly saturated and diffusion again slows. The shape of this curve has been found to be remarkably similar for many different types of innovations.

point in their lives. If gall stones block the bile duct, patients can experience severe abdominal discomfort and pain, heartburn, and indigestion. Prior to 1990, gall stones were treated in an open surgical procedure to remove the gall bladder. At this time,


Biomedical Engineering for Global Health

SOS Village Presentation: June 29, 2007 Christina Lesotho I have to admit that I found myself the night before last feeling a bit nervous about rolling out the activities and interacting on an educational level with the students. Although I have worked with youth before and have had similar experiences with refugees from various African countries, I was still unsure of how the students would respond to our teaching styles and activities. I now know what some teachers describe as the butterflies in their stomach the night before their very first day of teaching. After setting up our projector in the community hall, students began trickling in, one by one. We introduced ourselves, and again I was embarrassed by my lack of ability to pronounce their names. The initial ice was broken when they asked if I had music on my computer. They requested hip-hop or R&B and knew a few Usher songs and a Will Smith song I played. We kicked off the first day with an ice breaker game in which the children split into two groups and lined up behind a blanket. When the blanket dropped, the students had to shout the name of their classmate on the other side of the blanket as fast as possible. It was so funny to watch their reactions as the blanket dropped and to keep them from peeking underneath the blanket. I had my moment as the strict teacher and did my share of nagging. It turns out the concept of a pre-test is not common around here and even when we told them they could leave their names off the test, we had the toughest time convincing them that the pre-test we were giving out was only for our own knowledge – that it would never be graded or shown to their teachers. They were whispering to each other and trying to share answers left and right. I definitely did my part to stand over the mischievous ones, pace back and forth to scare them a bit, and laughed inside as their little eyes turned quickly away from their neighbors’ papers as they looked up and saw me coming by. I am really not sure how much they liked the powerpoint style of presentation, but I think it was something different and kept their attention fairly well. They were happy to get up in between slides and play the snowball game. We had them write down a question they were curious about relating to HIV, science, health, life in general and keep it to themselves. They then split up into two teams on opposite sides of the room and we told them to crumple up the papers and begin throwing them at the opposing team. It was absolute mayhem and turned out even better than we imagined. There was a question asking what love is. This age group was 12 years old and above, and this question puzzled me at first. Remembering that many of these children are orphans and know only the love of a house mother who takes care of nine others, I took back a bit of my confusion and began thinking of the millions of orphaned children who most likely have never known unconditional love. Nothing pains me more than thinking of the plight of orphaned children, and every day that I walk

Technology diffusion


through SOS, I am so thankful for the care and opportunities available to these children. I only wish this experience could be multiplied for so many others. We prepared this morning for our afternoon at SOS again by finding appropriate images to describe the male and female reproductive systems since we found ourselves motioning around our own bodies most of the time yesterday. It’s always such a pleasure to walk into the SOS Village because there are always children walking around who are so excited to see us and greet us with the warmest smiles. I want to get to know every child I pass, and I wish I had time to spend with each one of them. Today’s lesson and activities also went well, and the children are gaining so much knowledge that I think the questions I asked during some of the games were too easy. No one really got a kick out of the Twister board we brought, though, and they were all too shy to be the ones to play on the board. Overall, we found the boys to be more open and curious than the girls. The girls had questions and would whisper them to their most vocal friend, but the questions were few and I would really like the opportunity to sit in a smaller group setting with the girls. They are all sweet and welcoming, and seem to want speak with us on a more intimate level. After the presentation and games, we got to run around with the younger ones a bit and catch them on the playground. Again, they loved posing for pictures and seeing themselves when the picture came up on the digital camera. I cannot explain the way the children look at me, but maybe a few of the shots can capture their spirit, personality, and love of life.

Stomach Pancreas

surgical removal of the gall bladder was the most common non-obstetric surgical procedure in many countries. While effective and relatively safe (mortality rate 0.3–1.5%), this procedure required that patients stay in the hospital for about a week, and most lost 30 days of

Gall bladder

time from work. Laparoscopic cholecystectomy was a new procedure developed in the 1980s which allowed for shorter hospitalization, more rapid recovery, an earlier return to

Common bile duct

Pancreatic duct Duodenum


Figure 14.3. The gall bladder is located just beneath the liver. It stores bile made in the liver. Upon contraction, bile is secreted into the small intestine to aid in digestion. When gall stones form in the gall bladder, they can block the outflow of bile, leading to acute inflammation which must be treated surgically. SEER Training Modules, Pancreatic & Biliary Cancer, US National Institutes of Health, National Cancer Institute 2009.

work, and significant financial savings. In this procedure (Figure 14.4), the patient receives anesthesia, but only a small incision is made at the navel enabling a thin fiber-optic video camera to be inserted. The surgeon then fills the abdomen with carbon dioxide gas. Two needle-like instruments are inserted into the space created by filling the abdomen with gas. These instruments can be seen on the video monitor and serve as tiny hands, used to pick up the gall bladder and to manipulate the intestines. Using such instruments, the surgeon clips the gall bladder artery and bile duct, and safely dissects and removes the gall bladder and gall stones. The gall bladder is then removed


Biomedical Engineering for Global Health

Figure 14.4. In laparoscopic cholecystectomy, the abdomen is filled with CO2 gas to create a space for surgery. A camera inserted through the navel is used to visualize internal organs and small surgical instruments are used to manipulate tissues. Arteries and ducts leading to the gall bladder are clipped so that the gall bladder can be removed. When resected, the gall bladder is removed through the incision in the navel. c 2008 Nucleus Medical Art, Illustration  all rights reserved.

through the navel incision. The entire procedure takes 30–60 minutes. It requires only three puncture wounds which do not need sutures. The procedure does not

eral days, they can resume normal activities. The rate of complications for the laparoscopic procedure is about the same as for standard gall bladder surgery. Compli-

result in a scar; the three puncture wounds leave very slight blemishes and the navel incision is barely visible. Laparoscopic cholecystectomy has been called the most significant major surgical advance of the 1980s. This procedure was the forerunner of a new era of minimally invasive surgery. The benefits of the less invasive procedure include the ease of recovery, since there

cations include nausea and vomiting which may occur after the surgery. In addition, injury to the bile ducts, blood vessels, or intestine can occur, requiring corrective surgery. In about 5 to 10% of cases, the gall bladder cannot be safely removed by laparoscopy. In these cases, standard open abdominal surgery is then immediately performed. Let’s examine the rate of diffusion for this new

is no incision pain as occurs with standard abdominal surgery. Following laparoscopic cholecystectomy, about 90% of patients can go home the same day. Within sev-

surgery. In fact, no technique in modern times has become so popular as rapidly as laparoscopic cholecystectomy. Its diffusion in healthcare is unprece-

Technology diffusion


The tools of laparoscopic surgery The figures in this box show (a) an insufflator, (b) a trocar, (c) a camera, (d) laparoscopic cholecystectomy (surgery to remove the gall bladder), and (e) laparoscope with illuminator and camera system. Copyright 2009 by the University of Georgia, College of Veterinary Medicine. All rights reserved. (a)




(e) Monitor S-VHS or BNC Connection Cables

23 Light Cable Light Source Laparoscope Camera Head

dented. The technique was introduced in 1989; Figure 14.5 shows the percentage of gall bladder removal surgeries which were performed laparoscopically in the subsequent years. By 1992, 50% of all cholecystectomies in Medicare populations and 75% to 80% of all cholecystectomies in younger populations were performed through the laparoscope. Today, the laparoscopic procedure is the most widely used treatment for gall stone disease. Laparoscopic cholecystectomy resulted in a 22% decrease in the operative mortality


Video Signal Processor


rate for cholecystectomy [4]. In fact, the introduction of this new, minimally invasive procedure led to a substantial increase in the overall rate of cholecystectomy, as patients who were previously not good candidates for an open surgery could tolerate the risks of the minimally invasive procedure. Given the success and rapid diffusion of this procedure, how was its innovator viewed? The man largely responsible for the innovations that enabled laparoscopic surgery is Kurt Semm. Semm, a gynecologist,


Biomedical Engineering for Global Health was a gynecologist, he was accused of having “surgeon envy,” and was accused of trying to enter into general surgery to bolster his “operation ego.” Later, reflecting back Semm said, “Both surgeons and gynecologists were angry with me. All my initial attempts to publish on laparoscopic appendectomy were refused with the comment that such nonsense does not and will never belong to general surgery.” As we know today, Semm displayed an ability to push his ideas through despite skepticism and suspicion. In fact, without Semm, the laparoscopic revolution may have been postponed by many years. Can we use our growing knowledge about the determinants of technology diffusion, to speed the adoption

Figure 14.5. Rate of diffusion of laparoscopic cholecystectomy [4].

of new health technologies which are proven to be safe and effective? A number of strategies have been suggested [1], including the following.

contributed more than 80 medical device inventions during his life. They included the electronic insuffla-

1. Find sound innovations. In the next chapter, we will consider how medical research is funded, and how

tor, the use of thermocoagulation, the loop ligator, and methods to suture structures during laparoscopy [5]. Semm’s brother and father owned a medical instru-

this impacts the development and diffusion of new innovations. 2. Find and support innovators. As the story of Kurt

ment company which rapidly produced instruments for him that he could test. His ability to rapidly prototype and test new instruments enabled Semm to attempt to perform increasingly complex procedures endoscopi-

Semm indicates, innovators who lead major changes often come from outside the field, and their contributions are not always initially appreciated. 3. Invest in early adopters. Finding leaders within the

cally. Semm developed minimally invasive approaches to perform surgical procedures in both the areas of gyne-

field to champion a new change can decrease the resistance to that change. The speed of diffusion can

cology and general surgery. In 1985, Semm’s techniques were used to perform the world’s first laparoscopic appendectomy. At the time, the laparoscopic approach was said to reduce the problem of adhesions which frequently formed following open surgeries. How did Semm’s peers react to these

be increased by providing avenues to connect innovators and early adopters. 4. Make the activity of early adopters more visible. In making decisions about whether to adopt new innovations, the early majority looks to the activity of early adopters. Providing formal opportunities for

advances? Semm was said to have gone “absolutely crazy.” He was asked to undergo a brain scan by his colleagues. His lectures were initially greeted with laughter and derision. His techniques were initially viewed as too

these groups to interact can help increase the speed of diffusion. 5. Trust and enable reinventions. Often early adopters and the early majority adopt new innovations only in

expensive and too dangerous. Further, colleagues said that Semm exaggerated the problems of adhesions associated with open surgeries. At the time, most surgeons saw no reason to change a well established working method into a complex technical manner. Because he

part, adapting them to make them most effective in their local setting. Innovators are sometimes resistant to these changes, believing that they are an indication of resistance. Supporting effective local reinvention can speed diffusion.

Technology diffusion 6. Create room for change. Adoption of new innovations requires energy; if people are not given time and resources to support this, new innovations cannot diffuse. 7. Lead by example. Adoption of new innovations requires change at all levels in a system; leaders themselves must be open to change if diffusion is to occur. New innovations arise and are proven effective through a combination of basic science research, engineering design and clinical trials. In the next chapter, we will consider how health-related research is funded and regulated. We will see that the ways in which research is financed and regulated have an enormous impact on both innovation and the diffusion of innovation. This provides a unique opportunity to modify funding and regulatory policies to implement and reinforce the seven lessons considered above.

Homework 1. Contrast the rate of diffusion of the following two innovations: vitamin C to prevent scurvy in sailors, and laparoscopic cholecystectomy. 2. How are innovators, such as Kurt Semm, often viewed by others when introducing new


technologies for health problems that replace treatments considered by leaders in the field as already successful? 3. Why does it take so long for a promising new technology to reach the market in the United States? What hurdles must researchers overcome to deliver an innovation that is marketable?

References [1] Berwick DM. Disseminating innovations in health care. Jama. 2003 Apr 16; 289(15): 1969–75. [2] Balas E, Boren S. Managing clinical knowledge for health care improvement. In: Bemmel J, McCray A, eds. Yearbook of Medical Informatics. Stuttgart, Germany: Schattauer; 2000. [3] Papageorgiou C, Savvides A, Zachariadis M. International Medical Technology Diffusion. Journal of International Economics. 2007; 72(2): 409–27. [4] Ferreira MR, Bennett RL, Gilman SC, Mathewson S, Bennett CL. Diffusion of laparoscopic cholecystectomy in the Veterans Affairs health care system, 1991–1995. Effective Clinical Practice. 1999 Mar–Apr; 2(2): 49–55. [5] Litynski GS. Kurt Semm and the fight against skepticism: Endoscopic haemosis, laparoscopic appendectomy and the Semm’s impact on the ‘laparoscopic revolution.’ JSLS. 1998; 2: 309–13.

15 Regulation of healthcare technologies

As technologies diffuse from the research laboratory into clinical practice, we frequently encounter a tension between the goal of ensuring that products are fully tested before they are made available to the general public and the desire to make potentially life-saving technologies rapidly available to patients in need. In the United States, the Food and Drug Administration (FDA) is responsible for ensuring that new health technologies made available to patients are safe and effective. The FDA regulates the manufacture, testing and sales of chemical and biological agents and medical devices; its jurisdiction encompasses a range of products that together account for one-fourth of all consumer spending in the USA [1]. In this chapter we will consider the process that the FDA follows in considering whether to approve new drugs and medical devices. We will see that this process has evolved over time – over the past hundred years, the role of the FDA has shifted from a system in which a company could market a drug unless the FDA could prove that it was unsafe to one where drug manufacturers must first obtain permission from the FDA at nearly every step in the testing, production and marketing process [1]. This increasing burden of regulation has occurred largely in response to tragedies which could have been prevented. Yet, we will see that, even today,

Figure 15.1. Six out of ten Americans take one or more dietary supplements a day (multi-vitamins, amino acids, weight-loss cures, and herbal tonics) [3].

FDA regulation of dietary supplements is substantially less stringent than that of drugs, sometimes making it difficult for the FDA to act and prevent harm before it occurs. The distinction is important because more than 50% of Americans use some form of dietary supplements, including vitamins, minerals, amino acids, enzymes, herbs and other botanicals (Figure 15.1) [2]. In 2004, Americans spent more than $19 billion on dietary supplements [3]. We have already seen the important health

Regulation of healthcare technologies benefits of dietary supplements. At the middle of the eighteenth century, scurvy killed more British sailors than war; simply supplementing the sailors’ diet with



vitamin C prevented scurvy [3]. Adequate dietary folic acid early in pregnancy can prevent neural tube defects in the fetus. Calcium builds strong bones and can prevent osteoporosis [2, 3]; low levels of vitamin B12 may lead to dementia [3]. Yet, unlike drugs and medical devices, the FDA does not review the safety and effectiveness of dietary supplements before they are marketed [4]. The FDA can only act to prohibit sale of a dietary supplement if it can prove that it presents a risk of injury [4]. The dangers of this approach are illustrated by the recent ban on sales of the dietary supplement ephedra (Figure 15.2), following the death of Steve Bechler, the 23 year old pitcher for the Baltimore Orioles [2, 3]. Bechler died in February, 2003 of heatstroke after taking an over-the-counter product containing ephedra. When it was sold, ephedra was the most popular supplement in USA, generating sales of more than $1B/year, and accounting for approximately 10% of the annual sales of the US supplement industry [3]. Nearly ten years before Bechler’s death, the FDA began to hear reports of the risks associated with ephedra. Yet, in 1999, it was estimated that more than 12 million people in the USA were using ephedra [2]. The risks of ephedra use (particularly when used with caffeine), include an increased risk of heart attack, stroke, palpitations, anxiety, psychosis, and death [3]. There is no evidence that ephedra containing supplements are effective except for short term weight loss [5]. Although ephedra products accounted for less than 1% of all dietary supplement sales, a study in 2003 found that they accounted for 64% of all adverse events associated with dietary supplements [4]. The rate of strokes was found to be statistically higher in users taking doses of ephedra above 32 mg a day; many ephedra containing supplement labels recommended that patients take more than 100 mg of ephedra per day [4]. In February of 2003, the FDA sent warning letters to manufacturers of ephedra challenging them to remove unproven claims on their product labels [2]. In February of 2004, the FDA moved to ban sales of ephedra because it posed an unreasonable


Figure 15.2. (a) The dietary supplement ephedra is a naturally occurring substance derived from plants [4]. Wilhelm Thom´e. 1885. ¨ Flora von Deutschland, Osterreich und der Schweiz. Gera, Gemany. (b) The primary active ingredient in ephedra is ephedrine. Synthetic ephedrine is found in many over-the-counter remedies and some prescription drugs for treating colds, asthma, and nasal congestion [5], including Sudafed. Although the two products contain the same active ingredient (albeit at very different dosages), they are subject to a very different regulatory process in the USA.


Biomedical Engineering for Global Health

risk to patient safety. The ban took effect in April of 2004 [5]. In contrast to dietary supplements, the FDA requires

make it through this entire process. For every 5000– 10,000 drugs that enter pre-clinical testing, on average only ONE makes it to market. It has been estimated that

that manufacturers provide scientific evidence to indicate that new drugs are both safe and effective before they can be marketed. Generally, data from both preclinical and clinical trials are required to gain approval to market a new drug. Pre-clinical testing in cells and animals must be completed to assess potential toxicity before FDA approval for human studies is granted. In order to test a new drug in people, a manufacturer must

the cost of developing one new drug today ranges from $0.8 billion–$1.7 billion [6]. Do the current regulations governing approval for new drugs effectively balance protection from risk associated with potentially unsafe new drugs against gaining access to potentially effective new drugs? Are the fundamentally different approaches to the current regulation of drugs and dietary supplements appropriate? To

obtain an Investigational New Drug (IND) approval from the FDA. With an IND, human clinical trials are then allowed. In phase I trials, the safety of a compound is assessed by administering low doses of the new drug to a small group of healthy volunteers (20–100 volun-

help answer these important questions, it is helpful to examine the history of how the FDA regulates drugs, dietary supplements and medical devices. In contrast to the current complex regulatory process for drugs, in the early 1900s, the manufacture and mar-

teers). If successful, then phase II trials to assess the

keting of drugs was much less closely regulated. In the

effectiveness of a compound can commence, usually in 100–300 patients who suffer from the condition targeted by the drug. The final step before seeking FDA approval to market a new drug is to carry out a randomized phase

early 1900s, the patent medicine business accounted for more newspaper ads (Figure 15.4) than any other kind of product, a situation that is eerily similar to today’s widespread Internet advertising for dietary supplements

III clinical trial. After conducting phase I, II, and III clinical trials, a manufacturer can file an NDA (New Drug Application) for permission to market the new drug. If

[3]. At the time, many supplements contained substantial amounts of alcohol; others were even laced with cocaine, caffeine, opium and morphine [3]. Manufactur-

the NDA is approved, the manufacturer must still carry out studies to monitor for unanticipated complications of the drug and to study the longer term effects of drug exposure. This is called post-market surveillance. Dur-

ers were not required to disclose the contents of patent medicines on the product label. Largely in reaction to the sickening conditions of the meat-packing industry described in The Jungle by

ing this period, any adverse effects observed must be reported to the FDA [1]. Figure 15.3 illustrates the current drug development and approval process in the USA. The process is expensive, time consuming and complex. Not many drugs

Upton Sinclair (Figure 15.5), Congress passed the Pure Food and Drug Act in 1906 [7]. This law permitted the newly formed Bureau of Chemistry, precursor to the FDA, to ensure that labels on foods and drugs contained no false or misleading advertising; in addition, labels were required to contain accurate levels of 11 dangerous ingredients including alcohol, heroin and cocaine [7]. While the Food and Drug Act provided some protection for consumers, it did not require that companies obtain approval before marketing new drugs. The FDA could act only after harm had occurred. In the early 1930s, sulfanilamide was used as an antibiotic to treat streptococcal infections. It had been used safely as a

Figure 15.3. Current drug development and approval process in the USA. Pharmaceutical Research and Manufacturers of America.

pill for years; to help make the drug easier to use for children who have difficulty swallowing pills,

Regulation of healthcare technologies


Day 2 at SOS: Sex-ed and answering Snowball questions: July 3, 2007 Sophie Lesotho On our second day at SOS, we incorporated a sex-ed lesson into our presentation. We explained a bit about the female and male reproductive systems using anatomy cartoon visuals that we found online. The kids were very curious and asked many many questions. We inevitably had to deal with much giggling during our session, which I guess is normal. I remember having sex-education taught to me in 5th grade and thinking it was the funniest thing in the world. We also focused on answering questions that came up the day before during our Snowball Activity. At the beginning of the activity we asked everyone to write down any question they had about health, HIV/AIDS, sex, etc., that they would normally be embarrassed to ask their teachers or ask in front of their peers. The purpose of this activity was to decrease the stigma associated with HIV . . . and just to answer any questions they had in a friendly and non-judgmental environment. After everyone was done writing their questions, we split them up into two teams and had them crumple up the pieces of paper they had written on into “snowballs.” At the count of three, we had a huge snowball fight! This part was fun – the kids got really into this and were screaming and laughing. We felt that it was important to answer the questions that the kids had on this day since we came across many reoccurring questions on HIV/AIDS such as: – – – – –

What is the difference between HIV and AIDS? How does HIV attack the body? What are ARVs? What is a white blood cell? Can HIV be cured?

And we came across other serious questions such as: – If a girl is living with her brother and gets raped by him, what will happen? – What is love? – We were told at school that condoms are not 100% safe. Why are they not and what are we supposed to do? Later in the session, we played a version of Jeopardy asking the kids different questions covering the topics of immunity, Global AIDS, Prevention/Transmission, and Treatment. The kids participated well in this game and seemed to enjoy it very much. We found quickly that the kids knew too much for the questions we had made up for our game! They were answering almost all of them correctly.

Massengill, a Tennessee company found they could dis-

Elixir Sulfanilamide, they experienced severe abdomi-

solve the drug in diethylene glycol (antifreeze) [7, 8]. They tested their formulation for flavor, appearance, and fragrance, but not for toxicity [3]. At the time, safety tests were not required by law [8]. The new product was called Elixir Sulfanilamide; it was shipped

nal pain, nausea, vomiting, and convulsions [3]. One woman who lost a child to the tragedy wrote to President Franklin Delano Roosevelt: “Even the memory of her is mixed with sorrow for we can see her little body tossing to and fro and hear that little voice scream-

all over the country, and within weeks, more than 100 people were dead, most children [7]. After taking

ing with pain and it seems as if it would drive me insane” [3].


Biomedical Engineering for Global Health

Figure 15.5. Upton Sinclair on the cover of Time Magazine. Associated Press.

another tragedy to change this provision. In the 1960s, the William Merrell company hoped to market the drug Figure 15.4. Lydia Pinkham’s Vegetable Compound was an example of a patent medicine advertised in the newspaper. Pinkham’s compound was advertised as “a positive cure for all those painful complaints and weaknesses so common to our female population.” In 1914, the American Medical Association analyzed Pinkham’s compound and found it to contain 20% pure alcohol and 80% pure vegetable extracts [3]. Schlesinger Liberary, Radcliffe Institute, Harvard University.

thalidomide in the USA. It submitted an NDA to the FDA to market the drug [8, 9]. However, an FDA medical officer, Frances Kelsey (Figure 15.6), refused to

In reaction to the tragedy, in 1938, Congress passed

mide caused severe birth defects, and the FDA quickly acted to seize the supply of thalidomide that had been distributed. In response to this averted tragedy, President Kennedy signed the Drug Amendments Act into law

the Food, Drug and Cosmetic Act (FD&C Act) [3, 7, 8]. The FD&C Act gave the FDA the authority it needed to regulate such products before they came on the market. It required that companies test new drugs for safety before marketing. Companies were required to submit test results to the FDA in a New Drug Application (NDA). At this point in time, companies could test drugs in patients without approval from the FDA. It took yet

approve the NDA because of insufficient data regarding the safety of the drug. Even though the drug was never approved to be marketed in the USA, the company distributed over two million tablets in the USA for investigational use, which was not regulated by the FDA at that time. By 1962, it became clear that thalido-

in 1962. The new law required that FDA be provided with full details of planned clinical investigations of new drugs, requiring previous animal testing before initiating human clinical trials. In addition, it required that manufacturers prove new drugs to be both safe

Regulation of healthcare technologies


developing heart disease, diabetes, and cancer. In 1984, Kellogg’s, together with the National Cancer Institute, launched a marketing campaign for All-Bran cereal that illustrated how a low-fat, high-fiber diet might reduce risk of certain cancers [3].

Research in alternative medicine The National Center for Complementary and Alternative Medicine (NCCAM) is the branch of the NIH which oversees scientific research on complementary and alternative medicine (CAM). The mission of NCCAM is to rigorously and scientifically explore complementary and alternative healing practices, to train complementary and alternative medicine researchers, and to disseminate accurate information to the public and professionals.

Figure 15.6. Frances Kelsey received the President’s Distinguished Federal Civilian Service Award in 1962, from President John F. Kennedy; she was the second woman ever to receive the award [8, 9]. Her actions at the FDA are credited with preventing thousands of birth defects in the United States. Courtesy of the National Library of Medicine, NIH.

and effective before marketing them. The FDA was given control of all prescription drug advertising; previously this had been regulated by the Federal Trade Commission. While the FDA is also charged with regulating food safety, up until the 1960s, the line between foods and drugs was fairly clear and it was usually straightforward to determine whether a new product should be regulated as a food or a drug. However, over the past few decades, the line dividing foods and drugs has become increasingly less clear [3]. For example, in the 1970s, several government commissions issued recommendations encouraging Americans to alter their diets if they wanted to have longer, healthier lives [3]. We have seen that changes in diet can reduce a patient’s risk of

Over this same period, the popularity of dietary supplements increased dramatically. Between 1990 and 1997, the use of herbal remedies increased almost fourfold in the USA [10]. In 1994, after intense lobbying from the dietary supplement industry, Congress passed the Dietary Supplement Health Education Act (DSHEA). DSHEA permitted supplement manufacturers to make statements about the role of their products in health without FDA approval, provided that they could substantiate the claims with scientific evidence and that they included a disclaimer that the statements had not been evaluated by the FDA [7]. Unlike drug manufacturers, companies selling supplements are not required to prove products are effective or safe before marketing them [3]. Table 15.1 outlines the types of claims which can be made regarding dietary supplements under DSHEA, and whether FDA review of supporting data and approval are required [10]. Manufacturers are not permitted to make claims linking their product to a specific disease without FDA review. For example, without FDA review, manufacturers CANNOT say that a product reduces cholesterol but they CAN say it maintains healthy cholesterol levels if they have

Challenges of health technology regulation in developing countries Many developing countries struggle with the challenges of effectively regulating foods, drugs and medical devices. As commerce becomes increasingly global, these challenges are a concern for patients and consumers everywhere. The recent deaths of more than 100 people in Panama due to cold medicines and antihistamines containing diethylene glycol are eerily reminiscent of the 1938 Elixir Sulfanilamide tragedy in the United States. The Panamanian tragedy resulted when the government of Panama manufactured cold medicines and antihistamine syrups with what it believed was pharmaceutical grade glycerin imported from Barcelona. On its way to Panama, the solvent passed through three trading companies on three continents, originating in China. Unfortunately, the barrels of solvent sent to Panama did not contain pharmaceutical grade glycerin, which costs about $1815/ton; instead the Chinese manufacturer had substituted the less expensive solvent diethylene glycol which is similar in appearance to glycerin, but costs only $725–845/ton. The company was not certified to make pharmaceutical grade ingredients, and falsely certified the purity of the solvent. After Panamanian children began to die as a result of taking the cough syrup, Panama asked the CDC to test the medicine and found that it contained the poison diethylene glycol. The figures in this box show the original source of the solvent in China, and tracks its c 2007 New York Times Graphics). shipment and labeling as it made its way to Panama (copyright  (a)




As a result of the tragedy, the FDA has now recommended that all glycerin shipments be tested for diethylene glycol. More recently, in 2007 the FDA issued a consumer advisory warning people to discard all toothpaste made in China after federal officials discovered toothpaste containing diethylene glycol in Miami, Los Angeles, and Puerto Rico. Diethylene glycol was not listed as an ingredient on the label of the tainted toothpaste, but was discovered after FDA officials began to test Chinese made toothpaste following similar discoveries in toothpaste shipped from China to Latin America [19].

Regulation of healthcare technologies


Table 15.1. Claims that can be made about dietary supplements, according to the FDA. From [10].  c 2004 American Medical Association. All rights reserved.

Echinacea is one of the most commonly used cold remedies in USA. A recent clinical was carried out to determine whether echinacea could reduce the duration of the common cold. In a placebo controlled trial involving 400 children with common colds over four months, researchers found that the placebo worked just as well, and that children taking echinacea were more likely to develop a rash than those receiving placebo [3].

evidence to support this. They CANNOT say that echinacea cures disease, but with evidence CAN say it has natural antibiotic activities and is considered an excellent herb for infections of all kinds [3]. Do the majority of supplement retailers follow these rules? A recent study published in the Journal of the American Medical Association reviewed information presented about the eight most widely used herbal supplements found using the five most popular Internet search engines [10]. More than 80% of retail sites contained statements that made health-related claims (Table 15.2); more than half claimed to treat, prevent, diagnose or cure specific diseases even though these disease-related claims had not been reviewed by the FDA as required [10]. In addition, half of the sites with health-related claims omitted the required disclaimer that the statements had not been evaluated by the FDA [10].

Under DSHEA, there are almost no standards that regulate how dietary supplement pills are made, and they are not required to be tested once they are made [3]. DHSEA also places the burden of proving that a supplement is mislabeled on the FDA [7]. In response to the ban on ephedra, Congress has considered, but not passed, several bills that would modify 1994 law so that many unregulated botanical supplements would be treated more like drugs than like foods. How does the regulation of medical devices compare with that of drugs and dietary supplements? The FDA did not regulate medical devices before 1938. Similar to drugs, after the Food, Drug and Cosmetic Act was passed in 1938, the FDA could only challenge the sale of medical devices it believed were unsafe and could only remove them from the market after patient injuries had occurred as a result of their use. We have seen the rapid innovations in medical technology that occurred throughout the 1960s; in the field of cardiology alone, the heart – lung machine, prosthetic heart valves and many other technologies first came into use. The FDA initially tried to regulate many medical devices as drugs. After a number of catastrophic failures of life sustaining medical devices in the 1970s, including some types of heart valves and pacemakers, there was broad recognition that different rules were needed to regulate devices [11]. In 1976, Congress passed the Device Amendments to FD&C Act. These amendments recognized that no single regulatory policy would work for all devices. The rules for regulating the manufacturing and sales of a


Biomedical Engineering for Global Health

Table 15.2. Illegal health-related claims about dietary supplements are commonly found on retail websites.

tongue depressor should be different from those which are applied to an artificial heart.

Manufacturers must show that a new device is both safe and effective and receive approval from the FDA

The 1976 Amendments recognized three classes of medical devices. Class I devices pose the least risk to patient; they are not life sustaining, and encompass about 30% of medical devices including devices such

prior to marketing the device [1, 11, 12]. Because medical devices were not formally regulated until 1976, there are two paths to obtain approval to market a new device.

as X-ray film, tongue depressors, and stethoscopes. The FDA requires that manufacturers use a process called Good Manufacturing Practices when making some Class I devices (stethoscopes, but not tongue depressors); essentially this entails keeping extensive records about the source of suppliers for all parts of the device. Class II devices are not life sustaining, but must meet performance standards. Class II devices make up about 60% of medical devices and include things like blood pressure monitors, and catheter guide wires. Class III devices are for use in supporting or sustaining human life, and they pose the greatest risk to patient. About 10% of medical devices are considered to be class III devices, including stents, heart valves, and LVADs. These devices face the most stringent approval process [1, 11, 12].

Pre-market Notification Process (510K) The 510K is the approval path for new class I or II devices which are considered to be substantially equivalent to a device already on the market prior to 1976. The manufacturer notifies the FDA 90 days prior to when they plan to market the new device through a 510K application. In order to be considered substantially equivalent, a new device must have the same indications for use, and be no more risky and no less effective than a pre-1976 device [1].

Pre-market Approval Application (PMAA) The PMAA is the approval path for all new class III devices, as well as the path for class I or II devices that

Regulation of healthcare technologies


History of regulations The history of regulating drugs in the USA is the repeating story of misfortune, disaster, and tragedy – leading to reforms in drug regulation 1906: Pure Food and Drug Act Drug labels could not contain any statement regarding therapeutic effect which is false and/or fraudulent. The FDA could act only after drugs were marketed [6]. To prevent marketing of an ineffective drug, it was not enough to show that the product did not work; the government had to show that the seller knew the claims it made were false [6]. 1938: Food, Drug and Cosmetic Act Marked the beginning of era in which it is illegal to market a new drug without FDA approval. New drugs could not be marketed without first notifying the FDA and allowing agency time to assess safety. The seller’s belief regarding the product’s value was no longer relevant. The central issue became does the product really work [6]? 1962: Drug Amendments to FD&C Act Converted pre-market notification system into pre-market approval system where the FDA must review evidence of drug safety and effectiveness. Required that evidence of safety and efficacy come from well-controlled investigations by qualified experts. The FDA has the authority to prevent harm before it occurs [6]. 1976: Device Amendments to FD&C Act Devices are assigned to one of three classes. Based on the class of the device, the FDA may require pre-market approval or simply provide oversight of the manufacturing process and device labeling [6]. 1994: Dietary Supplement Health & Education Act Deregulated the dietary supplement industry. Manufacturers are not required to notify the FDA before they market supplements and they are not required to test safety or efficacy. Manufacturers are allowed to make health-related claims without FDA review if: they are supported by scientific evidence, they do not mention a specific disease, and are accompanied by a disclaimer that indicates statements have not been evaluated by the FDA [9, 10].

have no pre-1976 equivalent. Manufacturers submit a PMAA and provide the FDA with full reports of studies to show the safety and efficacy of the new device. All investigational studies must have been carried out with prior FDA approval under an IDE [1].

In the device approval process, the FDA considers the device and its intended use together. After the man-

An IDE is required for clinical trials using an investigational device which will be used in a way that could

ufacturer submits a request for marketing approval, the FDA convenes an advisory panel to act on the request. The advisory panel includes physicians and scientists with expertise in the field of use relevant to the device. In addition, the panel includes two non-voting members: one consumer representative and one industry representative. Although the FDA is not required to follow the recommendations of panel, it usually

pose a significant risk to study participants [11, 12]. An IDE application includes a complete description of the device as well as the planned study.

does [12]. Once a device is approved for use, a Medical Device Reporting System is used to detect device related

Investigational Device Exemption (IDE)


Biomedical Engineering for Global Health

Humanitarian use exemption

Recently approved devices

In order to provide an incentive for companies to develop devices for rare conditions, the FDA provides a third path for device approval. If a

Information about newly approved medical devices can be found on the FDA’s website. cfdocs/cfTopic/MDA/mda-list.cfm?list=1

device is designed to treat or diagnose a condition that affects fewer than 4000 patients/year, would not otherwise be available without exemption, and no comparable device is available, companies can apply for approval under a humanitarian use exemption. To obtain approval, the manufacturer must show that patients will not be exposed to unreasonable or significant risk of injury or illness by device.

For example, on October 24, 2003, the agency approved a new type of coronary artery stent , the NIRflex Stent System, to treat coronary artery atherosclerosis. A summary of the action and the evidence of the safety and efficacy of the device can be found at: The approval letter from the FDA specifies the conditions of the approval, and a copy of the letter can be found at:

problems in a timely manner. Clinicians are required to report serious injuries or deaths that may have been caused by or related to a medical device within ten days. The FDA requires post-marketing surveillance for both new devices and drugs. In 2004–2005, two new drugs were removed from the market due to side effects identified after they had been approved for use. The case of the selective non-steroidal antiinflammatory drugs, Celebrex, Bextra and Vioxx, provides a good example of both the advantages and challenges of the regulatory process [13–16]. Vioxx was manufactured by Merck, and was introduced in 1999 as an alternative to the available non-selective, nonsteroidal, anti-inflammatory drugs (NSAIDs) for treatment of osteoarthritis related pain [17]. At the time, the existing non-selective NSAIDs were effective in treating this pain, but these drugs produced gastrointestinal side effects such as ulcers. These existing nonspecific NSAIDs inhibited two types of COX enzymes; it was thought that inhibition of COX 2 was responsible for the anti-inflammatory effect (and pain reduction), while inhibition of COX 1 caused the gastrointestinal side effects. Vioxx selectively inhibited COX 2, so it was thought that it would relieve pain without gastrointestinal side effects [18]. Vioxx received

approval from the FDA in 1999, and sales grew rapidly [18]. A 1996 study sponsored by Merck showed that COX 2 inhibition could also precipitate the formation of thrombus in the vascular endothelium [18], which suggested that the drug might have cardiovascular side effects. Despite the biological plausibility of this hypothesis, none of the clinical studies that were part of the 1998

Regulation of healthcare technologies


Grassroots Soccer week: July 19, 2007 Sophie Lesotho We have been collaborating with an organization called Grassroots Soccer, which mixes HIV education with sports activities. A coordinator for Grassroots Soccer in Lesotho, Refiloe, helped us all of last week to conduct the different activities with the children. One of the best games was “HIV Dodge.” The activity clearly shows the role of the immune system, how HIV attacks it, and how ARVs attack HIV. It is a dodgeball-like game where all of the people standing around the human in a circle act as germs to attack the human in the middle with a dodgeball. The human cannot move, but the immune system acts as a defense to prevent the human from getting hit. The children then count how many times the human is hit. In the next round, HIV comes in and holds the hands of the immune system behind his/her back. The human is hit many more times than when HIV was absent. In the final round, ARVs come in and hold HIV’s hands behind his/her back and the human is once again protected against the attack of germs. All of the games were cleverly thought out and fun to participate in. What bothered me though was that the kids didn’t seem to remember the messages as much as they did the games themselves. This somewhat surprised me because I had the notion the games would always work to drive in the educational messages. Unfortunately, this does not always seem to be the case. Overall however, the kids participated well in the activities and I hope that the activities helped to increase their understanding of HIV concepts. At the end of the week, we gave the kids post-test evaluations to see how much they learned during the weeks we have spent with them (we gave them pre-test evaluations during our first session). I also interviewed some of the children on what they have learned about HIV/AIDS, what their favorite activities were, and also what their aspirations are for the future. We will continue working with the SOS children over the next two weeks that we are here, but we will be reiterating what we have already taught them – except this time, their role will be to teach others. We are having the kids perform and put together “Immunity Skits” that illustrate the role of the immune system, how HIV affects it, and how ARVs work (much like the dodgeball game except using acting). We will record these skits on the video camera and hopefully show this tape in the clinic waiting room and in schools back in Houston.

NDA for Vioxx were designed to evaluate cardiovascular risk associated with the drug [17]. Subsequent to its initial approval, a large study of 8000 patients was carried out to expand the possible indications for use of Vioxx [17, 18]. In late 1999, the

DSMB monitoring the study noted a 79% greater risk of cardiovascular death or serious cardiovascular event for those taking Vioxx [18]. The board recommended that the study continue, but that a plan to monitor adverse cardiac events be developed. A number of the


Biomedical Engineering for Global Health

members of the DSMB had consulting relationships with Merck, and it is unclear whether this potential conflict of interest influenced their recommendation [17].

the events in each trial were examined by an independent, external committee or not [18]. As a result, many have recommended reforms to the drug devel-

When data from the study were finally published, the study authors used different stopping dates for the cardiovascular adverse events than they did for the primary study outcome. This highly irregular procedure was not described in the methods section of the paper. More importantly, it had the effect of underestimating the risk of cardiovascular adverse events [17]. In 2002, the FDA required that the Vioxx label be changed to indi-

opment and regulatory process. For example, it has been suggested that academics engaged in industrydesigned studies should have full access to clinical data and that these data should be made accessible to the public [17]. After Merck withdrew Vioxx from the market, the FDA began a systematic reexamination of the risks associated with both selective and non-selective NSAIDs.

cate that physicians should exercise caution in prescribing Vioxx for patients with a history of ischemic heart disease [18]. Another large study in 2004 reported that Vioxx was associated with increased cardiovascular risk, but only

In 2005, FDA asked Pfizer to withdraw the selective NSAID Bextra from the market because the riskbenefit ratio was judged to be unfavorable [16]. The FDA advisory panel reviewing the data from Bextra voted 17 in favor, 13 against keeping Bextra on the

after 18 months of use [17]. Subsequently, the study

market; but FDA judged that, given the closely split

authors, who included five Merck employees and other authors who had received consulting fees from Merck, noted that their study contained flaws in the statistical analysis which understated the risk of cardiovascular

decision, regulatory action should be taken [13]. The advisory committee assessing the risk of Celebrex unanimously recommended that it remain on the market; FDA concurred and recommended labeling changes to

side effects [17]. Not long afterwards, two other randomized studies noted the increased cardiovascular risk of Vioxx [18],

highlight the increased cardiovascular risk associated with the drug. Finally, FDA also recommended labeling changes for non-prescription non-selective NSAIDs

causing the FDA to issue a cautionary note warning of the risk of cardiovascular events associated with the drug, and requiring a black box warning to be added to the drug label. Subsequently, Vioxx was voluntarily

(e.g. Advil, Motrin) to highlight the potential cardiovascular and gastrointestinal risks associated with these products. The story of Vioxx illustrates an important challenge

withdrawn from the market by Merck [17], but only after more than 80 million patients had taken the drug [18]. While the continued surveillance of Vioxx did ultimately identify the increased risk and lead to FDA action, the action may have been delayed as a result of conflicts of interest. Because Merck had billions of dollars at stake, and also paid consulting fees to some

in the development of new health technologies. The current development path for new medical products has become increasingly challenging and costly [6]. Just over the last decade, the costs of developing new drugs have escalated rapidly (Figure 15.7) to more than $1.7 billion. In part, late failures of new drugs due to unexpected adverse effects contribute to these increasing product development costs.

of the academic investigators who carried out the clinical studies, published the results, and served on the DSMB of the trials, ethical concerns were raised [17, 18]. In fact, in later reviewing the adverse cardiovascular event data from 18 different clinical trials of Vioxx, investigators showed that the only source of variation

The increasing costs of development have been paralleled by increases in health-related research and development (R&D) funding in the United States. Figure 15.8 shows that, over the period 1993–2003, R&D funding increased dramatically, both in the private sector and from the federal government primarily via the

in whether Vioxx was judged to be associated with increased risk of myocardial infarction was whether

National Institutes of Health. Despite this increase in R&D spending, there has been a dramatic decline in

Regulation of healthcare technologies 2.0

(a) Investment required for one successful drug launch (discovery through launch)

$1.7B Launch


300 US Pharmaceutical R&D Spending

Phase III/File


$1.1B Launch

1.0 Critical Path

Phase II

Phase III/File Phase II

Critical Path

Phase I Preclinical

Phase I Preclinical

0.5 Discovery





Figure 15.7. The costs of developing one new drug have rapidly escalated [6]. Rebuilding Big Pharma’s Business Model In Vivo. c 2008. Windhover Information Inc. This article cannot Nov. 2003  be reprinted without express permission of Windhover Information.

Indexed Growth (1993 = 100)

$ Billions


Total NIH Budget





0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year

same time period (Figure 15.8). If the cost and complexity of developing new health technologies continues to increase in a way that outpaces R&D investments, health innovation will likely decline [6]. In the final part of this chapter, we examine how health-related research is funded in the USA and what actions are being taken to sustain and expand innovation. Figure 15.9 provides a detailed picture of who carries out research in the USA and who pays for it. In this graph, research is divided into three types: basic research, applied research and development. The graph on the right illustrates that most basic research is carried out in universities and colleges. As we move to applied research and development, we see that most is carried

Percent 100 Source of funds

80 70

Total NMEs Rec’d by FDA Original BLSs

60 50 40 30

20 10 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Year

Figure 15.8a,b. Despite an increase in R&D spending, the number of applications for new drugs and biologic agents has declined dramatically over the past decade [6].

Percent 100 Performing sector

Billions of constant 1996 dollars 160 Development

120 60



Submissions to FDA

the number of new applications submitted to the FDA for approval of new drugs and biological agents over the

Applied research

40 0 1953 1960


Basic research 1970











0 Development

Federal Government

Applied research Industry

Basic research

Universities and colleges

FFRDCs = Federally Funded Research and Development Centers

Development Other nonprofit

Applied research


Basic research

Figure 15.9. Source of funds and performing sector for US research and development in 2000. Courtesy of FDA.


Biomedical Engineering for Global Health

out in private industry. The graph on the left shows who pays for R&D; basic research is largely supported by the federal government, whereas applied research and development is funded primarily by industry. The federal government funds approximately 36% of all medical research in the USA. This is mostly funded through the National Institutes of Health which is organized into a number of different disease- and technology-related institutes. In 2006, the annual NIH budget was $28 billion. The NIH budget doubled from 1998 to 2003, and has been flat through 2007. While the focus of the NIH is on basic research, the contributions to development of new products has been substantial. A US Senate Report in May 2000 examined the role of public funding in the development of the 21 new drugs introduced between 1965 and 1992 considered by experts to have had highest therapeutic impact on society. The report found that public funding of research was key in developing 15 of the 21 drugs. Three of these drugs, captopril (Capoten), fluoxetine (Prozac), and acyclovir (Zovirax), had more than $1 billion in sales in 1994 and 1995. Others, including AZT, acyclovir, fluconazole (Diflucan), foscarnet (Foscavir), and ketoconazole (Nizoral), had NIH funding and research to help in clinical trials. The NIH carries out internal research and also makes grants to academic and industrial researchers to carry out health related research. In general, the NIH sets research priorities by issuing requests for proposals through one of its institutes. Investigators write a proposal, describing their hypothesis, the significance of the proposed research, any preliminary results they have obtained, their planned research design and methods, and the plans to protect any animals or human subjects to be used. The NIH convenes a panel of experts in the field who critique the proposal and provide a numeric score to be used in determining whether the work should be funded. Following the review, the investigator receives both the score and the written comments. Each institute within the NIH then reviews the scores of all submitted proposals and decides which should receive funding. The scores range from a best of 100 and a worst of 500; typically proposals scored

National Institutes of Health

r National Cancer Institute r National Eye Institute r National Heart, Lung, & Blood Institute r National Human Genome Research Institute r National Institute on Aging r National Institute on Alcohol Abuse & Alcoholism r National Institute of Allergy & Infectious Diseases r National Institute of Arthritis and Musculoskeletal and Skin Diseases r National Institute of Biomedical Imaging & Bioengineering r National Institute of Child Health & Human Development r National Institute on Deafness & Other Communication Disorders r National Institute of Dental & Craniofacial Research r National Institute of Diabetes and Digestive and Kidney Diseases r National Institute on Drug Abuse r National Institute of Environmental Health Sciences r National Institute of General Medical Sciences r National Institute of Neurological Disorders & Stroke r National Institute of Nursing Research r National Library of Medicine

in the range of 100–170 are funded. Approximately 10–15% of submitted proposals are funded, although the fraction of funded proposals has been dropping steadily over the last four years because the NIH budget has not kept pace with the number of submitted proposals. What does the future of biomedical research and technology diffusion hold? Today, many researchers and policy makers are concerned that the converging challenges of flat research budgets and the increasing complexity of new technologies and their regulation will limit our ability to achieve the potential health gains that

Regulation of healthcare technologies could result from new basic science knowledge. Both the NIH and the FDA have introduced new programs to address these challenges. Recognizing the increasing time required for new technologies to diffuse from the bench to the bedside, in 2003 the NIH announced a series of new initiatives called the Roadmap to increase the emphasis on translational research. These initiatives are focused on creating new research approaches and teams that can work to speed the transition from the research laboratory to the clinical environment [6]. In parallel, to address the challenges associated with the increasingly complicated product development path, the FDA announced the Critical Path Initiative [6]. The goal of this initiative is to develop new scientific and technical methods to improve the predictability and efficiency on the path from laboratory prototype to commercial product [6]. For example, one pharmaceutical company has estimated that drugs which initially appeared promising but later failed due to liver toxicity observed in clinical trials have cost more than $2 billion in the past decade [6]. New predictive biological or computational models to predict liver failure early in the drug development process could have an enormous impact to speed and reduce the cost of drug development. These may include animal or computer methods, or biomarkers to predict risk of future toxicity, or new clinical evaluation techniques. Figure 15.10 illustrates the process of product development, from basic research to FDA approval, and illustrates where improvements in translational research and the critical path can help speed the successful development and diffusion of new health technologies.

Figure 15.10. The process of developing new health technologies and its relation to translational research and the critical path [6]. Adapted from: Innovation or Stagnation – FDA paper.


Missing our SOS Kids Already: July 27, 2007 Sophie Lesotho I am in love . . . with a little girl named Lisemelo. She is the most beautiful and sweetest little girl I have ever met in my life. Every time we walk by her home at the SOS orphanage, she is always waiting on the side of the road for us because she knows by now the times that we usually walk by. She just stands there and smiles at you so sweetly. Then, when you run up to her to give her a hug, she opens her arms so wide for you to pick her up (she is a tiny two-year-old). I wanted to cry today though . . . instead of walking back to the clinic from the SOS Village, Christina and I were picked up. As we drove by little Lisemelo’s home, she was standing out on the street and she could see us in the car as we approached. I waved to her and her face looked so sad when she realized that there would be no hugs today. I looked out the back car window after we had passed and I could see her walking up closer to the street and looking after our car. I was so sad . . . oh goodness, what do I do? I am so in love with little Lisemelo! But! I love all of our SOS kids. They are so funny. I caught a recording of them doing a dance-off of sorts in front of the camera today. I was filming while Christina was trying to teach the kids a hip-hop dance that they are learning as part of their Immunology Skit which they will be performing next week. Their awesome personalities shined through so well in this recording. I wish that there was some way to post it up on the blog. As I filmed, all of the kids (from ages four all the way up to 16) were laughing and jumping in front of the camera showing off their best moves. They kept pushing each other aside to hog the camera view and were laughing as they tried to steal the spotlight from each other. It is one of the funniest and cutest things I have ever seen. I love our SOS kids . . . I don’t know what I’m going to do without them. Christina teaching the SOS kids a dance:


Biomedical Engineering for Global Health

Homework 1. You have developed a new type gold nanoparticle to improve a physician’s ability to detect cancer at the earliest possible stages. Many Internet based health food stores sell solutions of identical gold nanoparticles to “lift your body’s performance, and fight off germs, viruses, bacteria, allergens, pathogens and pollution.” Yet, it will be several years before you can begin clinical trials to determine whether this same nanoparticle can improve early cancer detection. a. Describe the current differences in FDA regulation of this particle when it is used as a dietary supplement vs. as a drug. b. Briefly summarize the history of government regulation of drugs in the USA, noting the year of major changes in legislation and the primary changes in regulation associated with these laws. 2. Consider the differences in regulation of drugs and dietary supplements. a. If I wish to market a new dietary supplement which I claim will improve immune function, am I obligated to provide scientific data to the FDA indicating that it is safe and/or effective before I can sell it? b. If I wish to market a new drug to treat pancreatic cancer, am I obligated to provide scientific data to the FDA indicating that it is safe and/or effective before I can sell it? c. The following is an actual Internet ad for an herbal supplement. If you were employed by the FDA to monitor and investigate these ads please indicate statements you see that may not follow legal guidelines in the United States. Explain why you chose each of your selections. If you did not identify anything, explain why. There’s a new nutritional supplement available for people suffering from Type 2 diabetes. It’s called The Body Rejuvenator, marketed by Lafayette Miracle Solutions. It contains two key ingredients – green tea extract and cinnamon. The first thing to realize is that nutritional supplements can very successfully control blood sugar in diabetics. Both

green tea and cinnamon are well-known to help control blood sugar so that you don’t have such wild blood sugar swings (and potentially don’t need as much insulin either). Also, there are many other benefits documented from taking both green tea and cinnamon. Green tea is noted for its anti-cancer effects, as well as its ability to aid in weight loss, which is something that diabetics are typically concerned with. 3. I wish to market a new dietary supplement. Indicate which of the statements in the list below I am legally allowed to put on the product label. If a statement would not be allowed, indicate why not. a. Acidophilus, Bifidus & Bulfaricus promote the health of the digestive tract. b. Black Currant Oil contains essential fatty acids that provide dietary support for normal healthy blood lipids and helps to support the cardiovascular system. c. SkinAnswer, a glycoalkaloid skin cream, as a treatment for skin cancer. d. Ephedra-free Total Leantm helps dieters increase their metabolism and boost their energy. e. MGN-3, a rice-bran extract, a treatment for HIV, the virus that causes AIDS. f. ZantrexTM-3 promises 546% more weight loss than the leading ephedra-based diet pill and that’s a fact. Here’s another fact: Zantrex-3 is way beyond ephedra, way beyond fat-burners, way beyond everything on the market today. Zantrex-3 is a new category of bifurcated weight loss compounds providing both rapid weight loss and incredible energy combined into a single power-packed Super Pill. New Zantrex-3 is so powerful you won’t find it in some Wal-Mart next to some “Flintrock” vitamin for kiddies. g. BeneFin, which is produced from shark cartilage, is a treatment for cancer. 4. Over the past 100 years, the role of the FDA in regulating drugs has changed significantly. Briefly describe the history of these changes. Contrast changes over time in the history of FDA regulation of medical devices. 5. Read the following article. Explain how, after passing all the safeguards of pre-clinical testing,

Regulation of healthcare technologies phases I–III of clinical trials, and required scrutiny of an FDA panel, problems such as these could occur. April 23, 2004 F.D.A. Seeks Reports of Stent Problems By GINA KOLATA The Food and Drug Administration is actively seeking reports of possible problems with a stent that came on the market last month, saying it has heard of serious medical complications in some cases. But Dr. Daniel G. Schultz, director of the agency’s office of device evaluation, said in a telephone interview on Tuesday that it was too soon to say whether there was a problem with the stent and, if so, what was causing it and what advice to give doctors and patients. The F.D.A. knows of 20 to 25 incidents, Dr. Schultz said, but the reports range from sketchy to highly detailed. “We’re fairly early in the process of assessing the reports”, Dr. Schultz said. “At this stage, our main goal is to gather more information.” The device’s maker, Boston Scientific, says that its stent is safe and is performing excellently and that any problems are extraordinarily rare. Paul LaViolette, senior vice president at the company, said more than 70,000 of the stents had been used in the United States since the device went on sale in March. “We have to conclude, and I will say this with a lot of experience, that this product is performing extremely well,” Mr. LaViolette said. But a few cardiologists reported in telephone interviews that they got into trouble after the stent, a small wire tube used to hold open arteries, was slipped into place. Like all stents, Boston Scientific’s stent, the Taxus Express2, comes packaged with a deflated balloon inside. A cardiologist threads the stent with its balloon into an artery. When the site of the blockage is reached, the doctor inflates the balloon, pressing the stent against the artery wall. Then the balloon is deflated and the catheter and


balloon withdrawn, leaving the stent flush against the artery, holding the vessel open. Some doctors said the balloon stuck on the stent when they were removing it. Some were able to free the balloon; some were not. Dr. William Campbell, director of the cardiac catheterization laboratory at Borgess Heart Institute in Kalamazoo, Mich., said a patient was rushed into emergency open heart surgery to remove the balloon and stent. Others, like Dr. Alejandro Prieto of Michigan State University, said that the balloon did not deflate and that he had to use a sharp wire to pop it. But then he also punctured the patient’s artery. “Those are serious problems,” said Dr. Schultz, who said the F.D.A. had received similar reports. Dr. Andrew Carter of Providence St. Vincent’s Medical Center in Portland, Ore., said that while his medical center had used several hundred taxus stents without incident he was nonetheless worried. “I have never had a balloon that did not deflate, or a device entrapment,” he said, referring to balloons that got stuck on stents. “You would never expect to see it. Period.” c 2004. The From the New York Times, April 23 

New York Times. All rights reserved. Used by Permission and protected by the Copyright Laws of the United States. The printing, copying redistribution, or retransmission of the Material without express writer permission is prohibited. 6. In 1937, a drug manufacturer attempted to modify sulfanilimide, an antibiotic for streptococcal infections, so that it was easier for children to take. Sulfanilimide had been used safely as a pill for years; however, most children can’t swallow pills. A company in Tennessee found they could dissolve the drug in ethylene glycol (antifreeze). The company tested their new solution for flavor, appearance, fragrance, but NOT for toxicity. They proceeded to ship it all over the country. Within weeks, many children had died. a. Was this legal at the time? b. How and when were federal laws reformed to prevent this from happening in the future?


Biomedical Engineering for Global Health

7. Suppose we wish to track the progress of a promising new drug through all stages of development and testing. a. How long does it typically take for a promising new drug to go from the research laboratory to the market in the United States? b. Describe the phases of study researchers must go through to develop a new drug before it can be marketed. For those phases of study which involve giving the drug to patients, give the typical number of patients involved and the goals of the clinical trial. c. What fraction of promising drug candidates actually make it to the market? d. What is the cost of developing a new drug today in the USA? e. Recently, several drugs to treat arthritis were withdrawn from the market or given a black box warning. Why were the problems with these drugs not discovered until after the FDA had approved their sale? 8. Contrast the role of the NIH and industry in providing funding to support medical research in the United States. 9. Read the article below and answer the following questions. Experiment: Closed-Heart Surgery Associated Press 16:30 PM Apr, 01, 2006. Used with permission from the Associated Press. c 2009. All rights reserved. Copyright  Dr. Samuel Lichtenstein cut a 2-inch hole between an elderly man’s ribs. Peering inside, he poked a pencil-sized wire up into the chest, piercing the bottom of the man’s heart. Within minutes, Bud Boyer would have a new heart valve – without having his chest cracked open. Call it closed-heart surgery. “I consider it some

lodged in just the right spot in the still-beating heart. The dramatic experiments, in a few hospitals in the United States, Canada and Europe, are designed to find easier ways to replace diseased heart valves that threaten the lives of tens of thousands of people every year. The experiments are starting with the aortic valve that is the heart’s key doorway to the body. The need for a less invasive alternative is great and growing. Already, about 50,000 people in the U.S. have open-heart surgery every year to replace the aortic valve. Surgeons saw the breastbone in half, stop the heart, cut out the old valve and sew in a new one. Even the best patients spend a week in the hospital and require two months or three months to recuperate.Thousands more are turned away, deemed too ill to survive that operation and out of options. Demand is poised to skyrocket as the baby boomers gray; the aortic valve is particularly vulnerable to rusting shut with age. The new experiments are a radical departure from that proven, if arduous, surgery. The artificial valves do not even look like valves, squished inside metal cages until they are wedged into place. Barely 150 of any type have been implanted worldwide, most in the last year. It is unclear if they will work as well as traditional valve replacements, which last decades. For now, the only patients who qualify for these valves are too sick to be good candidates for regular valve replacement. Some deaths during the earliest attempts at implanting the devices forced doctors to come up with safer techniques. Clinical trials apparently are back on track, and even the most skeptical cardiologists and heart surgeons are watching how these pioneers fare. The hope is that one day,

kind of magic,” said Boyer, who left the Vancouver, British Columbia, hospital a day later and was almost fully recovered in just two weeks. In Michigan, Dr. William O’Neill slipped an

replacing a heart valve could become almost an overnight procedure. “There’s lots of technical challenges that need to be overcome,” said Dr. Robert Bonow, a valve

artificial valve through an even tinier opening. He pushed the valve up a patient’s leg artery until it

specialist at Northwestern University, who is monitoring the research for the American Heart

Regulation of healthcare technologies


Association. “Most of us do think this is the future,” he said. O’Neill’s first successful patient in March

Paris-based CoreValve are testing versions of a collapsible valve made of animal tissue that is folded inside a stent, a mesh-like scaffolding

celebrated the one-year anniversary of his through-the-leg implant. “I call it a new birthday,” chuckled Fred Grande, 78, a Richmond, Michigan, car collector who took one of his beloved models for a fast spin less than a week after the procedure. “That’s the home run we want to hit with all the patients,” said O’Neill, cardiology chief at William

similar to those used to help unclog heart arteries. The difference is how doctors get the new valve to the right spot, pop open its metal casing and make it stick. The U.S. studies thread the Edwards valve through a leg artery up to the heart, known as “percutaneous valve replacement.” Unlike with open-heart surgery, doctors do not stop the

Beaumont Hospital in Royal Oak, Michigan. “It’s gratifying” to watch people once deemed beyond help bounce back, added Dr. Jeffrey Moses of New York-Presbyterian Hospital/Columbia University, who with O’Neill is leading the U.S.

patient’s heart. So the trickiest part is keeping regular blood flow from washing away the new valve before it is implanted. Once the device is almost in place, doctors speed the heartbeat until normal pumping pauses

study. One of Moses’ first patients is playing golf at

for mere seconds – and quickly push the new valve

age 92. The heart has four valves – one-way swinging doors that open and close with each heartbeat to ensure blood flows in the right direction. More

inside the old one. Inflating a balloon widens the metal stent to the size of a quarter, lodging it into place and unfolding the new valve inside, which immediately funnels the resuming blood flow.

than 5 million Americans have moderate to severe valve disease, where at least one valve does not work properly, usually the aortic or mitral valves.

So far, 19 Americans have been implanted this way, plus more than 80 other people worldwide, most of them in France by the procedure’s

Worldwide, roughly 225,000 valves are surgically replaced every year. Topping that list is the aortic valve. It can become so narrowed and stiff that patients’ hearts

inventor, Dr. Alain Cribier, and in Vancouver by Lichtenstein’s colleague, Dr. John Webb. Fourteen people in Canada, Germany and Austria have received the Edwards valve through

wear out trying harder and harder to push oxygen-rich blood out to the rest of the body. Calcium deposits accumulate on its tender leaflets. Touch one chipped out of a patient and it feels almost like a rock. With minimally invasive valve replacement, doctors do not remove that diseased valve. Instead, they prop it open and wedge an artificial

the ribs. That is a more direct route to the heart for patients whose leg arteries are too clogged to try the other experiment. Doctors make a tiny hole in the bottom of the heart muscle so the new valve can enter. Then they use the same balloon technique to wedge it inside the old valve. Talks have begun with the Food and Drug Administration about opening a similar U.S. study

one into that rigid doorway. “It’s ironic. You use the disease process to actually help hold your valve in place,” said Lichtenstein, of St. Paul’s Hospital in Vancouver, who helped create the between-the-ribs method.

later this year. CoreValve’s slightly different valve is being tested in Europe and Canada. It, too, is threaded up the leg artery. But it is made of pig tissue instead of horse tissue and has a self-expanding stent that requires no balloon. Doctors remove a

Edwards LifeSciences in Irvine, California, the biggest maker of artificial heart valves, and

sheath covering it and the stent’s metal alloy, warmed by the body, widens until it lodges tight


Biomedical Engineering for Global Health against the old, rocky valve. More than 45 have been implanted; CoreValve hopes to begin a U.S. study next year. Lead researcher Dr. Eberhard

he said. Eleven others have lived a year and counting. CoreValve reports five patients faring well a year later. Aside from those who did not

Grube of The Heart Center in Siegburg, Germany, expects within months to begin testing a newer version small enough to thread through an artery at the collarbone, another more direct route to the heart. The experiments come with some significant risks. Edwards temporarily halted the U.S. study last year after four of the first seven U.S. patients

survive the implantation, others have died from their advanced illnesses even though their new valve was working. It is the cases of astounding successes – Grande and Boyer, for example – that have other heart specialists taking note, Northwestern’s Bonow said. “Patients have to know what they’re getting

died. Initially, doctors threaded the valve up a leg vein, not an artery, a route that required tortuous turns inside the heart and sometimes damaged a second valve, O’Neill said. Twelve people have been implanted since the study restarted in

into,” he said. Many of the seriously ill are willing to chance the experimental procedure because “they’re so debilitated and . . . there have been some good examples of patients who have gotten better.” The bigger challenge, Bonow added, is

December using the artery route considered easier

whether to expand the studies to include less sick

and safer. All but one have survived and are faring well, researchers say. O’Neill and Moses – plus doctors at a third hospital, the Cleveland Clinic – have government

patients who could survive open-heart valve replacement but want to avoid its rigors. Already, there are such patients clamoring to be included. That is a difficult decision because even 80- and

permission to implant eight additional patients in the U.S. pilot study, which will be expanded if it goes well.

90-year-olds successfully can have regular valve replacement. When performed by the most skilled surgeons, risk of death from the operation is about

CoreValve’s first four patients died as doctors struggled to develop and learn the through-theartery technique, Grube said. For doctors, pushing the large valve through tiny, twisting arteries –

2 percent – but in less experienced hands, it can reach 15 percent, Bonow said. Just as using a balloon to unclog heart arteries is sometimes done on patients who would fare better

against regular blood flow and guided by X-rays – is laborious. Occasionally, they are not able to wedge it into position. Because they are squeezing a round valve into an irregular-shaped opening, there is a risk that the new valve will leak blood backward into the heart, also problematic. But once researchers master how to get the

with bypass surgery, researchers eventually will have to ask if patients would accept a less-thanperfect aortic valve if they could skip surgery’s pain and risks, said Dr. Michael Mack of Medical City Hospital in Dallas. “There is a trade-off, and how you make that trade-off is a totally gray area,” he said. But Vancouver’s Boyer, who had two previous

valve into place safely, the question becomes how much recipients benefit. Do these very ill patients live longer than expected? If not, does quality of life improve enough to warrant the procedure anyway? Three of French inventor Cribier’s original

open-heart surgeries for clogged arteries, said avoiding that kind of pain is not a trivial issue for patients. “They’re doing something to the field of medicine that’s going to make life a hell of a lot easier to people who’ve got that problem,” said a grateful Boyer, describing how he could

patients have lived 2 1/2 years so far, with a “return to normal life and no sign of heart failure,”

finally breathe easy after the through-the-ribs valve implant. “I think I’ll have a bunch of

Regulation of healthcare technologies other parts go bad before I have a problem with this.” a. Discuss the factors which are likely to affect the diffusion of this technology. Do these factors always benefit the patient? b. Why do you think the sample sizes are so low for the studies reported here? Consider what we learned in Chapter 12 about the trials of the AbioCor artificial heart. What factors do you think the FDA considers in decisions regarding the clinical trials reported here? 10. Consider a new implantable device that does not have any reasonably similar products already in the market. a. What class of device would the FDA consider this product? b. Would this device need a 501K or PMA approval process? c. Based on your answer to part b, describe the remaining steps needed to carry this product through to the market.

References [1] Merrill RA. Regulation of drugs and devices: an evolution. Health Affairs. 1994 May 1, 1994; 13(3): 47–69. [2] Administration USFaD. Evidence on the Safety and Effectiveness of Ephedra: Implications for Regulation. February 28, 2003. [3] Specter M. Miracle in a bottle. The New Yorker. 2004 February, 2004: 64–75. [4] Medicine NCfCaA. Reducing ephedra-related risks. 2003 [cited June 8, 2007]; Available from: http://nccam.nih. gov/health/alerts/ephedra/022803.htm [5] Rados C. Ephedra ban: no shortage of reasons. FDA Consumer. 2004 March–April 2004. [6] Administration USFaD. Innovation or stagnation?: Challenge and opportunity on the critcal path to new medical products. In: Services USDoHaH, ed. 2004. [7] Administration USDoFaD. History of the FDA. [cited 2007 June 8, 2007]; Available from: [8] Administration USFaD. A brief history of the Center for Drug Evaluation and Research. November 1997 [cited


2007 June 8, 2007]; Available from: [9] Administration USFaD. Thalidomide: important patient information. July 7, 2005 [cited 2007 June 9, 2007]; Available from: http://www.fda.goc/cder/news/thalidomide.htm [10] Morris CA, Avorn J. Internet marketing of herbal products. JAMA. 2003 September 17, 2003; 290(11): 1505–9. [11] Wholey MH, Haller JD. An introduction to the Food and Drug Administration and how it evaluates new devices: establishing safety and efficacy. Cardiovascular and Interventional Radiology. 1995 Mar–Apr 1995; 18(2): 72–6. [12] Pritchard WF, Jr., Carey RF. U.S. Food and Drug Administration and regulation of medical devices in radiology. Radiology. 1997 October 1, 1997; 205(1): 27–36. [13] Administration USFaD. FDA regulatory actions for the CoX-2 selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs): questions and answers. April 7, 2005 [cited 2007 June 9, 2007]; Available from: infopage/COX2/COX2qa.htm. [14] Administration USFaD. COX-2 selective (includes Bextra, Celebrex, and Vioxx) and non-selective non-steroidal anti-inflammatory drugs (NSAIDs). July 18, 2005 [cited 2007 June 9, 2007]; Available from: http://www.fda. gov/cder/drug/infopage/COX2/default.htm [15] Administration USDoFaD. FDA public health advisory: FDA announces important changes and additional warning for COX-2 and non-selective non-steroidal anti-inflammatory drugs (NSAIDs). April 7, 2005 [cited 2007 June 9, 2007]; Available from: [16] Administration USFaD. FDA News: FDA announces series of changes to the class of marketed non-steroidal anti-inflammatory drugs (NSAIDs). April 7, 2005 [cited June 9, 2007]; Available from: topics/news/2005/NEW01171.html [17] Krumholz HM, Ross JS, Presler AH, Egilman DS. What have we learnt from Vioxx? BMJ. 2007 January 20, 2007; 334(7585): 120–3. [18] Zarraga IGE, Schwarz ER. Coxibs and heart disease: what we have learned and what else we need to know. Journal of the American College of Cardiology. 2007 January 2, 2007; 49(1): 1–14. [19] Bogdanich W and Hooker J. From China to Panama, a trail of poisoned medicine. The New York Times, May 6, 2007.

16 Future of bioengineering and world health

As we conclude our study of bioengineering and world health, we stop to examine some of the health challenges that developing countries face and consider how efforts to develop new, appropriate technologies may help address these challenges. Ten million children under the age of five die every year throughout the world; 98% of these deaths occur in developing countries. This is more than twice the number of children born each year in the USA and Canada combined. It has been estimated that 2/3 of childhood deaths could be prevented today with available technologies feasible for low income countries. Yet, current technologies do not reach millions of children in need – in many cases because these technologies are presently too expensive, require infrastructure that is unavailable (e.g. refrigeration for heat labile vaccines), or cannot be delivered because of a lack of effective healthcare systems [1]. Advances in the biosciences, bioengineering, and public health are responsible for the dramatic gains in life expectancy achieved over the last century. Throughout this book, we have seen that these advances are not equally available to people throughout the world (Figure 16.1). A recent global checkup of human health observes “In far too many countries health conditions remain unacceptably – and unnecessarily – poor” [2].

Figure 16.1. In the pediatric ward at Kamuzu Central Hospital in Lilongwe, Malawi, there is one nurse on staff for every 80 pediatric patients. Owing to a shortage of staff, children cannot be admitted without a guardian to provide care for them. Because there are not enough beds, most patients must share beds, while their parents sleep on the floor beside them.

Despite these inequities, 90% of the $70 billion spent each year on health research and development is devoted to diseases that predominantly affect industrialized nations. It is clear that many of the technologies we have examined in this book – for example, left ventricular assist devices – are simply not available in many

Future of bioengineering and world health (a)


What stops life saving tools from reaching the world’s poorest children? The Disease Control Priorities Project notes “. . . health inequities have arisen largely from uneven adoption and implementation of health interventions associated with technical progress” and cites as priorities developing effective low cost interventions for neglected diseases as well as developing effective strategies to get interventions to neglected people [1].

The country of Swaziland, located in southern Africa, illustrates some of the challenges of HIV/AIDS in sub-Saharan Africa. Swaziland has a population of about one million people. In 2005, the HIV prevalence rate among adults was estimated to (b)

Figure 16.2. (a) Graph showing changes in life expectancy and per capita income for different countries throughout the world from 1975 to 1990. Progress in four sub-Saharan African countries is tracked over this period. Data from the USA are shown for comparison. (b) Graph showing these changes from 1975 to 2004. Note the dramatic drop in life expectancy due to the AIDS pandemic beginning in 1990. The free software tool at Gapminder World provides an excellent way to explore global changes in income and health metrics over time.

developing countries. At the same time, developing countries are facing a growing burden of cardiovascular disease. Furthermore, the impact of HIV/AIDS has slowed, and even reversed much of the progress to improve health in many African countries (Figure 16.2).

be a staggering 34%, the highest in the world. Life expectancy at birth in 2006 was only 41 years. More than 70,000 Swazi children have been orphaned as a result of HIV/AIDS [3].

In short, biotechnology and bioengineering research continues to transform the future of healthcare in developed countries, but ensuring that the benefits of research are available to all world citizens requires a new way of thinking, which must incorporate technology development as well as public policy and management of healthcare delivery. Often, due to limited infrastructure, it is not simply enough to provide existing technologies designed for use in developed nations; in many cases a new kind of technology – one which is robust and does not require disposable supplies – is


Biomedical Engineering for Global Health

needed to function effectively in the developing world (Figure 16.3). The growing importance of global health prob-

ural disasters all highlight the importance of increased efforts to address global health disparities. Thanks in large part to the substantial investments and efforts of

lems has received substantial attention lately. Concerns about worldwide spread of emerging infectious diseases, humanitarian desires for equity in access to healthcare, and the fragile balance of sustaining past advances in world health in the face of increasing conflict and nat-

the Gates Foundation, a number of high-profile scientific planning activities have delineated the research challenges most likely to lead to long term, substantial benefits in global health. These include the need for improved, cost effective point-of-care diagnostic

Departing thoughts: August 3, 2007 Yesterday, a young boy at the clinic lay paralyzed while several of the doctors struggled to keep him alive. His tiny body was on the examining table while beeping noises filled the room. The young boy could no longer breathe on his own. There are no child ventilators in this country. Why? None of the people we spoke to really knew for sure, but most speculated that this was due to a lack of organization of those who are in charge of inventory of medical supplies. The physicians decided to take the child to Bloemfontain, which is the town with the nearest South African hospital. I just thought about how this boy would have surely died if he was not in the care of the doctors at our clinic. The staff at the government hospital seems too overstretched to have ever taken the care to send the child to Bloemfontain where a simple ventilator could be obtained to save this child’s life. There is such needless dying and suffering in this country. And when I hear that many are dying due to carelessness or disorganization – something inside me just burns with anger. The country always runs out of CD4 reagents . . . As a result, I have heard it announced repeatedly throughout my stay here that there are no CD4 counts for the week, which is crucial for monitoring HIV patients and starting them on antiretroviral treatment. In a country where 1/4 of the population has HIV, having CD4 counts available is a necessity. Again, there are speculations that the CD4 reagents run out not because of a lack of funding, but instead because of people who are disorganized. Another thing that deeply angered and frustrated me today was seeing about 20 spotless Mercedes Benzs lined up at the airport along with several beautiful black Audis and about 20 shiny Toyota Camrys . . . The government here pays for these cars for their highest government officials, while the lower down government officials all get Camrys (but not to worry, starting next year all the Camrys get upgraded to Lexus cars). It is all so wasteful . . . Especially now, after the government has declared a state of emergency. This country suffered from its worst drought in 30 years – many were expected to starve to death or die from malnutrition this year. Why is the government paying for these fancy cars and not spending it on their people who are suffering and dying from a mere lack of food? The 500,000 rand or so spent on each of these cars could also easily be invested to send more kids to school. There is a lack of access to higher education for too many children of this country because they cannot afford to pay for high school tuition. Furthermore, while the local government hospital is in shambles, a beautiful new ministry building is being built right next to it. The conditions at the hospital are terrible – it is overcrowded, the staff is too few, the building is too old – there are even cockroaches crawling out of the children’s beds during the summer. There is absolutely no excuse for not making the rebuilding of the hospital a priority. People are dying needlessly due to the dreadful conditions of the place. But yet, a beautiful new ministry building is being built . . . and the rebuilding of the hospital will continue to be put off while the government officials keep driving their sparkling cars. . . .

Future of bioengineering and world health


devices, the need for improved vaccines to prevent infectious disease, and the need for novel delivery methods for vaccines and medications.

Lack of educational opportunity also contributes to poor health in the developing world. Young people with little or no schooling are up to twice as likely to contract HIV as those who have completed primary education. Comprehensive solutions to global health challenges must include strategies to expand access both to appropriate technologies and to education.

Figure 16.3. Another challenge of providing appropriate health technology in developing countries is the frequent lack of infrastructure, technical supplies and the difficulties associated with maintaining and repairing instrumentation. The photo, taken at a hospital in Swaziland, illustrates that the availability of technology does not always translate into the ability to use that technology to address health needs. The photo shows an X-ray imaging system in front of the hospital elevator. Both the elevator and the X-ray machine are broken, and parts to repair them could not be obtained. The X-ray machine is now used to block the entrance to the elevator.

Technologies for the developing world must adapt not only to inadequate resources, but also to unique economic, cultural, social, and environmental realities. Designing technologies like this is an exercise in extreme engineering, which can be approached through both novel high-tech and low-tech solutions. Efforts to design appropriate health technologies can improve care both in developing countries and in wealthy countries as

well. New, simple health tools that work in hospitals in developing countries will also perform in rural, remote, and underprivileged healthcare settings in the industrialized world. We have seen that in many developed countries, healthcare costs continue to escalate, partly due to progressively more complex and expensive technologies that offer only incremental health benefit. New, cost effective technologies have the potential to not only reduce inequities in healthcare, but also to reduce the costs of care for all.


Biomedical Engineering for Global Health

Low-tech solutions to health challenges in developing countries Queen Elizabeth Central Hospital is the main government hospital in Blantyre, Malawi. The neonatal intensive care unit at the hospital has only one commercially made neonatal incubator. Unfortunately, when the thermostat in the incubator broke, no spare parts were available to repair the incubator. Instead, Dr. Liz Molyneaux, chair of the Department of Pediatrics, and her colleagues invented their own incubator, which can be made for less than $100 using locally available materials. The Blantyre Hot Cot consists of a wooden crib, with a hinged Plexiglass cover. Four 60 W light bulbs which can be turned on independently are installed beneath the crib. The bulbs warm the air beneath the baby. Warm air rises up into the crib, and the temperature is controlled by adjusting the number of bulbs which are turned on. The neonatal intensive care unit contains 12 Hot Cot incubators; an excellent example of a low-tech solution which addresses the health challenge in a way that is affordable and can be operated in the current infrastructure.

Future of bioengineering and world health


Things I have seen: August 7, 2007 It has been difficult accepting the fact that my time is over here because there are few times when I have left a foreign place feeling at home. It is true that I have missed my family and close friends, but I am leaving feeling as if I have become acquainted with an incredible family at Baylor. The physicians, the staff, the patients, and the people I have come in contact with have been some of the most amazing individuals I have known, and I will never forget their hospitality and their welcoming spirits. I am forever grateful to the Beyond Traditional Borders program, BIPAI, our program directors and mentors who have created such an incredible opportunity and have allowed me to be part of such a wonderful mission. I have been told by many who have worked in developing settings or in some sort of volunteer work that those who attempt to teach others or contribute to a problem in some way end up leaving having learned much more than they could have ever taught others or end up gaining much more than they could have been left behind as some sort of contribution. I have found this to be true, and I have learned so much about the people, culture, health, education, challenges, and opportunities of this country in such a short period of time. I have felt such a unique combination of emotions all packed into a series of encounters and experiences that seem to blend into one another like one of the beautiful tapestries woven in rural villages here. It is as if all of life’s emotions can be packed into a single day’s work – happiness, frustration, empathy, anger, desperation, fulfillment. . . . I could go on and on. I am leaving with a refreshed and renewed perspective on global health and the complexities that exist when working on problems of such magnitude. I will miss the daily challenge of working on any aspect of HIV/AIDS and the tough questions I asked that ended up consuming my thoughts and conversations late into the nights. I have seen for myself the tragic truth that many speak of . . . of the needless deaths that occur daily and the completely preventable illnesses that young and innocent children die from. I have seen the “accidents of latitude” that Bono and Sachs speak of when they talk about the unthinkable disparities that exist among those who have been born in the developed world and those who have been born in areas such as sub-Saharan Africa. I have seen the struggling face of a baby who died of a simple case of diarrheal infection, and the face of her mother who thought her child was on her way to improvement. I have seen the determined faces of a medical team that went to great lengths to take a child to a South African hospital just to put a baby on a life-saving ventilator, a simple tool they lacked in this country. I have felt the pang of injustice, not injustice I have personally faced, but injustice that I have felt through my close encounters with children, mothers, grandmothers, health professionals, and people from all over the world working in this country. I have seen things and felt emotions that have left a lasting impression, and I only hope that I have been able to contribute a fraction of the impact I have felt myself and that I have been able to leave just one child with a fraction of the knowledge I have gained.


Biomedical Engineering for Global Health

Homework 1. The table shows the global disease burden and the R&D funding for several diseases. a. Define DALY. Why is the DALY a better measure of disease burden than mortality rate? b. Based on this chart, what recommendations would you make for future funding of R&D funding?

b. Explain the potential positive impact that each technology could have on health. c. Describe the criteria you used to select these technologies. 4. You are provided with $100 million dollars to make an investment in one of the three following areas. a. Launch of a public health program to educate people regarding the risks of lung cancer and the benefits of smoking cessation and which offers free smoking cessation programs. b. Development of a new screening test for early lung cancer which costs $100 and has a specificity of 95% and a sensitivity of 98%. c. Development of a novel pharmaceutical compound to treat lung cancer. The new treatment reduces side effects by improving the specificity of targeting and improves five year survival for lung cancer patients by 30%. Choose one of these options and explain your rationale for selecting this option. Be sure to justify why you feel the option that you selected

2. Over the past decade, describe the epidemiologic shift in burden of disease that has occurred in developing countries. Given this epidemiologic shift, recommend two specific changes that should be made to R&D funding priorities for chronic diseases. Discuss one challenge associated with each recommendation. 3. You are working for a non-profit organization dedicated to implement new health technologies in the world’s least developed countries. The board of directors wants to allocate their 2006 budget to the development of biotechnologies that will most improve health in developing countries. To encourage the successful application of these technologies to global health, they have requested that you conduct a study to determine which biotechnologies could have the largest positive impact on health in developing countries. Your final report must address the following questions. a. Mention at least three technologies that you would recommend for implementation.

is a better investment than the other two options. Please be as quantitative as possible in your justification. 5. Reflect on what we have learned about world health in relation to your own personal and career goals. How can YOU work to improve world health? Write down one personal goal illustrating how you will try to improve world health as a result of something you learned in this text.

References [1] Jones D, Steketee R, Black R, Bhutta Z, Morris S, and The Bellagio Child Survival Study Group. How many child deaths can we prevent this year? The Lancet. 362 (9377): 65–71, 2003. [2] Jamison DT, Breman JG, Measham AR, Alleyne G, Claeson M, Evans DB, Jha P, Mills A, Musgrove P, eds. Priorities in Health, The World Bank, Washington, DC, 2006. [3] World Health Organization, Key WHO Information,


3 by 5 initiative, 83 AbiCor artificial heart, 342–4 adjuvant, 199 administrative costs, 141–2 aging population, 136–7 AIDS, 79–85, 96–8, 105, 215 AIDSVAX, 218 anastamosis, 333 angina, 91–2, 327 angiogenesis, 258 angioplasty, 331–3, 334–5 antibiotic resistance, 67 antibody, 194–5 antigenic drift, 196 antigenic shift, 196, 210 antiviral drugs, 197 apheresis, 11, 12 Apligraf, 176 artemisinin, 75 artificial heart, 340 ARVs, 24, 27 assessment, 306 association, 39 atherosclerosis, 91, 171, 329, 334–5 avian flu, 198 Bacille Calmette Guerin (BCG), 96 bacteria, 187, 188 bacterial disease, 189

basophils, 187 B-cells, 199 BD SoloShot syringe, 214 bed net, insecticide treated, 75 Belmont Report, 232, 233–5, 378 bile, 369, 377 bioinstrumentation, 161, 162–9 Biojector2000, 214 biomaterials, 4, 172–4 biomechanics, 4, 169–72 biomedical engineering, 4, 152–3 biomedical imaging, 4, 153–62 biopsy, 285, 288–98 biosensors, 4, 161, 162–9 biosystems engineering, 4 birth asphyxia, 62 birth trauma, 62 blinded seroprevalence studies, 236–7 blood pressure, 324 bone marrow transplant, 11–16 brain–machine interface, 164 CA-125 blood test, 292, 293–4 cancer, 99–102, 104 cancer, breast cancer, 5–11, 13, 21, 169 cancer, burden, 248, 250 cancer, cervical cancer, 267, 268–81 cancer, early detection, 259 cancer, lung cancer, 104

cancer, mortality, 250–3 cancer, ovarian cancer, 285, 288–98 cancer, pain, 263–4 cancer, pathophysiology of cancer, 254 cancer, prostate, 25, 281–8 cancer, risk, 253–4 cancer, screening, 260 cardiac output, 325 cardiovascular disease, 90–4, 104, 320–49 cardiovascular disease, prevention, 347–9 CD4 lymphocytes, 80, 82, 215 cellular engineering, 178–81 cerebrovascular disease (stroke), 91, 93–4, 322 cervix, 268 chemotherapy, 10–11 childhood deaths, 77 chloroquine, 75 cholera, 32, 71–2 cirrhosis, 105, 106 clinical practice, 232 clinical trials, 14, 21, 23, 356–66 clinical trials, drugs, 361 Clinton plan, 145 clonal expansion, 195 cold chain, 212–13 colpcoscope, 161, 162–9



colposcopy, 272–3 columnar epithelial cells, 269 computer axial tomography (CT), 94, 157 congenital anomalies, 75–6 coronary arteriography (cardiac catheterization), 92, 93 coronary artery bypass grafting (CABG), 92, 328–31, 334–5 coronaviruses, 64 crash test, 89 data data, collection, 27 data, economic data, 40–3 data, health data, 32–40 data, primary data, 27 data, secondary data, 27 Data Safety and Monitoring Boards (DSMB), 363 Da Vinci robotically assisted coronary surgery system, 332 decision analysis, 24, 27 Declaration of Helsinki, 232 dehydration, 70–3 descriptive statistics, 358 design specification, 152 Device Amendments to FD&C Act, 383–4, 385 diagnosis, 25 diarrheal disease, 67–74 diastole, 323 Dietary Supplement Health Education Act (DSHEA), 381–3, 385 dietary supplements, 369, 376–8, 381–3 digestive diseases, 105–6 digital image analysis (DIA), 278 diptheria, 202 direct immunofluorescence assays (DFA), 65, 66 direct visual inspection with acetic acid, 315–16 directly observed treatment short-course (DOTS), 98 disability adjusted life years (DALY), 37, 106 discounted life expectancy, 315 Don Thomas, 11 Drug Amendments Act, 380, 385 drug delivery, 172–4

echocardiography, 326 economic evaluation, 305–10 edema, 336 edible vaccines, 201 effect size, 362 effectiveness, 26, 308 efficacy, 26 ejection fraction, 325 electrocardiagram, 92 engineering design, 151–2 enzyme-linked immunosorbents assays (ELISA), 65, 81, 82 eosinophils, 187 Ephedra, 369, 377 epidemiologist, 32 epidermal growth factor receptor (EGFR), 181 epithelial tissue, 269 epithelium, 254 experimental approach, 360 FDA, 203, 376–91 fetal alcohol syndrome, 76 fetal ultrasound, 61 flow cytometer, 165–6 folate deficiency, 76 Food Drug and Cosmetic Act, 380, 385 Freireich, Emil J., 119–21, 170–1 fusion inhibitors, 83 gall bladder, 369, 376–8 Gallo, Robert, 215 gallstones, 369 Gates Foundation, 107 Global Alliance for Vaccines and Immunization (GAVI), 214 global burden of disease, 55 Global Outbreak Alert and Action, 35 glucose-sodium transporter, 71 HDCT + BMT, 13–17, 21, 23–4 health expenditure, 136 Health Maintenance Organization (HMO), 121, 175 health-policy space, 306 health systems, 113, 115, 119 health systems, centrally planned economy, 127 health systems, comprehensive health system, 77, 115

health systems, entrepreneurial health system, 115 health systems, free market economy, 128–31 health systems, single payer system, 121–4 health systems, socialist health system, 115, 125–7 health systems, US health system, 119–21, 170–1 health systems, welfare-oriented health system, 115 healthcare reform, 143–6 heart attack, 92, 326–8 heart attack treatment, 328 heart failure, 335–7 heart failure, left ventricle, 336 heart failure, right ventricle, 337 heart failure, treatment, 337–8 heart rate, 325 heart transplant, 338–40 heart–lung machine, 155–6, 331 HeartMate III, 342, 348 Heckler, Margaret, 215 Helsinki Declaration, 234 hepatitis, 105 herd immunity, 204 high density lipoprotein (HDL), 323 highly active anti-retroviral therapy (HAART), 82–3 Hippocratic Oath, 234 histocompatibility markers, 12 HIV, human immunodeficiency virus, 28, 79–85, 105 HIV, prevention, 237–9 HIV, screening, 81 HIV, testing, 217 HIV, transmission, 237 HIV, vaccine, 215–22 HIV, vaccine, live attenuated, 220 HIV, vaccine, noninfectious, 218–20 HIV virus, 191 HIV, HIV-1, 218 HIV, HIV-2, 218 HPV, human papillomavirus, 270–1 HPV, DNA testing, 277, 278, 313–16 HPV, vaccine, 279 human development index (HDI), 41 hyperlipidemia, 91 hypernatremia, 73

Index hypertension, 91, 323 hypothesis, 151 hypothesis, generation, 358–60 hypothesis, testing, 360–1 immune system, 187, 198 immune system, adaptive immune system, 192, 194, 198 immune system, antibody-mediated immunity, 198 immune system, cell-mediated immunity, 194, 198 immune system, humoral immunity, 194 immune system, innate immune system, 192–4, 198 immune system, physical barrier, 192, 198 immunization, 202–3 immunoassay, 167–8 immunosuppressive drugs, 338 incidence, 34 infectious disease, 186–98 influenza, 190, 196–8 influenza, vaccine, 206, 209–11 Institute of Medicine (IOM), 29 Institutional Review Board (IRB), 235 International Health Regulations (IHR), 35 investigational new drug (IND), 378 iodine deficiency, 76 ischemia, 91 ischemic heart disease, 90–1, 93, 322 Jarvik-7 artificial heart, 342, 348 Jenner, Edward, 200–1 Jewish Chronic Disease Hospital Study, 231–2 jungle, the, 232, 378 Kaposi’s sarcoma, 33, 80, 81 Karp, Haskell, 341 King, Albert, 90–1 lab-on-a-chip, 166 Langer, Robert, 121, 175 laproscopic cholecystectomy, 371–2 lead-time bias, 281–4, 287 least developed country, 43

left ventricular assist devices, 344–7 liquid cytology, 275–6, 313–16 Littenberg, Benjamin, 22–4 low density lipoprotein (LDL), 323 lower respiratory infections, 63–7 low-income country, 42 lumpectomy, 10 lymphocytes, 187 lymphocytes, T-lymphocytes, 195 macrophage, 193 magnetic resonance imagin (MRI), 94, 157–8 major histocompatibility (MHC) complex, 195 malaria, 74–5 malaria, cerebral malaria, 75 mammography, 9, 267 Mantoux tuberculin skin test (purified protein derivative test), 96 mastectomy, 10 master console, 332 Medicaid, 119 medical device classification, 384 Medicare, 119 metastasis, 101 metastatic tumor, 258 methylmercury, 76 microfluidics middle-income country, 42 minimum clinically important difference, 363 MMR vaccine, 204–5 molecular engineering, 178–81 monocytes, 187 morbidity, 37, 106–7 Morbidity and Mortality Weekly Report (MMWR) mortality, 37, 55–7, 106–7 mortality rate, 37 15–44 years, 79–104 45–60 years, 104–6 birth–4 years, 57–79 infant, 37, 77–9 neonatal, 77 mother to child transmission, 85 MR angiography (MRA), 94 multi-drug resistant TB (MDR TB), 98 mutation, HIV, 197, 217–18 myocardial infarction, 92


nationally notifiable diseases, 34, 36 National Institutes of Health, 390–1 needle-free vaccine, 214 negative predicitive value, 266–8 neonatal period, 57 neutrophils, 187 new drug application (NDA), 378 NIH, 390 nuclear to cytoplasmic ratio (N/C ratio), 256 Nuremberg Code, 232, 234, 235 observational study, 360 oncogenic mutation, 257 open heart surgery, 328–31 optical microscope, 161–2 oral rehydration therapy, 70–3 Pap smear, automated, 276–7 Pap smear, Papanicoloau, 271–2, 273–5, 313–16 partograph, 63 Pasteur, Louis, 201 PATH, 63 pattern discovery, 296 per capita health expenditure, 40 per capita income, 39, 40 percutaneous transluminal coronary angioplasty (PTCA), 92 perinatal conditions, 60–3 perinatal period, 57 perspective, 308 pertussis, 64, 189, 190 PET, 158–60 physiology, 178 plaque, 92 plasma, 10 Plasmodium falciparum, 74 platelets, 10 Pneumocystis carinii pneumonia (PCP), 33, 80, 81 pneumonia, 64–7 pneumonia, bacterial, 64 pneumonia, viral, 64 point of service (POS), 121, 205–12 point prevalence, 35 polio, 202 positive predictive value, 266–8 pre-cancerous lesions, 256



preferred provider organization (PPO), 121 premature birth, 79 prescription drug costs, 137–8 prevalence, 35, 267, 268–81 prevention, 25 primary tumor, 258 priority rating, 143 prostate specific antigen (PSA), 284 prostatectomy, 285, 292–3 protease inhibitors, 82 purchasing power parity (PPP), 40 Pure Food and Drug Act, 378, 385 quality adjusted years of life (QALYs), 113, 306, 308 QuantiFERON-TB Gold (QFT-G) blood serum test, 96 quinine, 75 red blood cells, 10 regenerative medicine, 175–7 regulation of healthcare technologies, 376–91 rehabilitation, 25 rejection rates, 338–9 relative risk, 39, 40 research, 235 resistance, 83 resistant strains of bacteria, 67 restenosis, 174, 333 reverse transcriptase, 82 road traffic accidents, 85 RotaShield, 73 rotavirus, 67, 73–4 rubella, 76, 202 sample size, 361–6 San Antonio Contraceptive Study, 232 scientific method, 150–1 scientific misconduct, 16 screening, 25 scurvy, 368 self-inflicted injuries, 102–4 Semm, Kurt, 373–4

sensitivity, 264, 265–6 sensitivity analysis, 309 sepsis, 63 serum PSA test, 283 severe acute respiratory syndrome (SARS), 64 smallpox, 200–1, 202 smoking, 91 Snow, John, 32 sodium reabsorption, 70–3 specificity, 264, 265–6 sphygmomanometer, 324 squamous epithelial cells, 269 standard deviation, 358–60 standard of care, 26, 239–40 standardized sifference, 361, 362 Starling’s Law, 336 State Children’s Health Insurance Program (SCHIP), 119 stem cells, 11 stent, 172, 174, 333, 334–5 stent, drug eluting, 333 stroke, 93–4 ischemic stroke, 93 stroke volume, 325 suicide, 102 surgical cart, 332 syphilis, 231 systems biology, 178 systole, 323 technology assessment, 17, 21–2, 24–6, 29, 305 technology development, 4–5 technology diffusion, 368–75 technology utilization, 138–9 thalidomide, 33, 76, 380 ThinPrep, 275 thrombolysis, 94 tissue engineering, 175–7 training phase, 296 transcutaneous energy transmission, 343 transient ischemic attack, 93–4 treatment, 25 tuberculosis, 94–9

tumor tumor, benign, 99 tumor, malignant, 99 Tuskegee Syphilis Study, 229–31, 235 type I error, 361 type II error, 361, 362 ultrasound, 160–1 ultrasound imaging, 285, 292–3 Unicef, 72 uninsured, 142–3 unintentional injuries, 77, 85, 104–5 vaccination in developing countries, 212–15 vaccine, 199 vaccine, bivalent, 207 vaccine, carrier, 220 vaccine, DNA, 220 vaccine, inactivated organism vaccine, 199 vaccine, live attenuated vaccine, 199–200 vaccine, monovalent, 207 vaccine potency, 212 vaccine production, 121, 205–12 vaccine, subunit vaccine, 200 vaccine vial monitors (VVM), 213 variance, 358 Vioxx, 386–8 viral disease, 190 viral load, 80, 83 viral load, tests, 84 virus, 188 visual inspection with acetic acid (VIA), 277–8 Weekly Epidemiological Record, 34, 36 white blood cells, 10, 187 WHO, 32, 66, 72 Willowbrook Study, 231 World Food Programme (WFP), 39, 56–7 World Health Report, 34 X-ray, 157